From bd7597f632c4f1a57b8efdef4cc9fa470ae250e0 Mon Sep 17 00:00:00 2001 From: Maximilian Date: Wed, 15 Jan 2025 08:11:35 +0100 Subject: [PATCH] Tabpfn explainer (#302) --- CHANGELOG.md | 3 + .../tabular_notebooks/explaining_tabpfn.ipynb | 492 ++++++++++++++---- .../tabular_notebooks/tabpfn_values.npz | Bin 2151 -> 0 bytes .../tabpfn_values_explainer.npz | Bin 0 -> 2154 bytes .../tabular_notebooks/tabpfn_values_game.npz | Bin 0 -> 2155 bytes requirements.txt | 1 + shapiq/__init__.py | 8 +- shapiq/explainer/__init__.py | 3 +- shapiq/explainer/_base.py | 86 ++- shapiq/explainer/tabpfn.py | 120 +++++ shapiq/explainer/tabular.py | 74 ++- shapiq/explainer/tree/explainer.py | 11 +- shapiq/explainer/utils.py | 35 +- shapiq/games/__init__.py | 4 +- shapiq/games/imputer/__init__.py | 3 +- shapiq/games/imputer/base.py | 18 +- shapiq/games/imputer/tabpfn_imputer.py | 110 ++++ shapiq/interaction_values.py | 36 +- tests/conftest.py | 26 + tests/requirements/requirements.txt | 1 + tests/test_base_interaction_values.py | 12 + .../tests_explainer/test_explainer_models.py | 38 +- .../tests_explainer/test_explainer_tabular.py | 4 +- .../tests_explainer/test_tabpfn_explainer.py | 61 +++ tests/tests_imputer/test_tabpfn_imputer.py | 100 ++++ 25 files changed, 1067 insertions(+), 179 deletions(-) delete mode 100644 docs/source/notebooks/tabular_notebooks/tabpfn_values.npz create mode 100644 docs/source/notebooks/tabular_notebooks/tabpfn_values_explainer.npz create mode 100644 docs/source/notebooks/tabular_notebooks/tabpfn_values_game.npz create mode 100644 shapiq/explainer/tabpfn.py create mode 100644 shapiq/games/imputer/tabpfn_imputer.py create mode 100644 tests/tests_explainer/test_tabpfn_explainer.py create mode 100644 tests/tests_imputer/test_tabpfn_imputer.py diff --git a/CHANGELOG.md b/CHANGELOG.md index 0f2fd307..d66ff5de 100644 --- a/CHANGELOG.md +++ b/CHANGELOG.md @@ -1,6 +1,9 @@ ## Changelog ### v1.1.2 (2025-01-13) +- adds ``shapiq.TabPFNExplainer`` as a specialized version of the ``shapiq.TabularExplainer`` which offers a streamlined variant of the explainer for the TabPFN model [#301](https://github.com/mmschlk/shapiq/issues/301) +- handles ``explainer.explain()`` now through a common interface for all explainer classes which now need to implement a ``explain_function()`` method +- adds the baseline_value into the InteractionValues object's value storage for the ``()`` interaction if ``min_order=0`` (default usually) for all indices that are not ``SII```(SII has another baseline value) such that the values are efficient (sum up to the model prediction) without the awkward handling of the baseline_value attribute - renames ``game_fun`` parameter in ``shapiq.ExactComputer`` to ``game`` [#297](https://github.com/mmschlk/shapiq/issues/297) - adds a TabPFN example notebook to the documentation - removes warning when class_index is not provided in explainers [#298](https://github.com/mmschlk/shapiq/issues/298) diff --git a/docs/source/notebooks/tabular_notebooks/explaining_tabpfn.ipynb b/docs/source/notebooks/tabular_notebooks/explaining_tabpfn.ipynb index d9044cbb..57487985 100644 --- a/docs/source/notebooks/tabular_notebooks/explaining_tabpfn.ipynb +++ b/docs/source/notebooks/tabular_notebooks/explaining_tabpfn.ipynb @@ -28,12 +28,13 @@ "metadata": { "collapsed": true, "ExecuteTime": { - "end_time": "2025-01-10T13:55:35.932354Z", - "start_time": "2025-01-10T13:55:31.928667Z" + "end_time": "2025-01-14T16:27:18.169723Z", + "start_time": "2025-01-14T16:27:13.871771Z" } }, "source": [ "from importlib.metadata import version\n", + "import os\n", "\n", "import shapiq\n", "import tabpfn\n", @@ -50,8 +51,8 @@ "name": "stdout", "output_type": "stream", "text": [ - "shapiq version: 1.1.1.dev\n", - "tabpfn version: 2.0.1\n", + "shapiq version: 1.2.0\n", + "tabpfn version: 2.0.3\n", "Device: cpu\n" ] } @@ -70,8 +71,8 @@ { "metadata": { "ExecuteTime": { - "end_time": "2025-01-10T13:55:35.978513Z", - "start_time": "2025-01-10T13:55:35.933357Z" + "end_time": "2025-01-14T16:27:18.232240Z", + "start_time": "2025-01-14T16:27:18.173711Z" } }, "cell_type": "code", @@ -240,8 +241,8 @@ { "metadata": { "ExecuteTime": { - "end_time": "2025-01-10T13:55:35.994521Z", - "start_time": "2025-01-10T13:55:35.979512Z" + "end_time": "2025-01-14T16:27:18.248248Z", + "start_time": "2025-01-14T16:27:18.234238Z" } }, "cell_type": "code", @@ -283,8 +284,8 @@ { "metadata": { "ExecuteTime": { - "end_time": "2025-01-10T13:55:36.326775Z", - "start_time": "2025-01-10T13:55:35.995512Z" + "end_time": "2025-01-14T16:27:18.531271Z", + "start_time": "2025-01-14T16:27:18.250239Z" } }, "cell_type": "code", @@ -732,8 +733,8 @@ { "metadata": { "ExecuteTime": { - "end_time": "2025-01-10T13:57:53.128517Z", - "start_time": "2025-01-10T13:55:36.333769Z" + "end_time": "2025-01-14T16:28:12.553241Z", + "start_time": "2025-01-14T16:27:18.534261Z" } }, "cell_type": "code", @@ -749,8 +750,8 @@ { "metadata": { "ExecuteTime": { - "end_time": "2025-01-10T13:57:53.144447Z", - "start_time": "2025-01-10T13:57:53.129439Z" + "end_time": "2025-01-14T16:28:12.569239Z", + "start_time": "2025-01-14T16:28:12.554240Z" } }, "cell_type": "code", @@ -771,8 +772,8 @@ "name": "stdout", "output_type": "stream", "text": [ - "MSE: 0.27149947144257525 R2: 0.796390135236755\n", - "Average prediction: 2.0852828\n" + "MSE: 0.27150313127523246 R2: 0.7963873905609423\n", + "Average prediction: 2.0879457\n" ] } ], @@ -781,8 +782,8 @@ { "metadata": { "ExecuteTime": { - "end_time": "2025-01-10T13:57:53.331718Z", - "start_time": "2025-01-10T13:57:53.145436Z" + "end_time": "2025-01-14T16:28:12.854835Z", + "start_time": "2025-01-14T16:28:12.571243Z" } }, "cell_type": "code", @@ -811,15 +812,18 @@ "source": [ "## Explain TabPFN with shapiq\n", "Now that we see how TabPFN performs, we can use shapiq to explain the predictions.\n", - "First, we will use the KernelSHAP method to explain the predictions." + "\n", + "This notebook will now cover two different strategies to explain TabPFN:\n", + "1. **Remove-and-Contextualize**: This strategy removes features from the model and re-contextualizes the model with the new data points.\n", + "2. **Remove-and-Impute**: This strategy removes features from the model and imputes the removed features with the mean/mode of the training data." ], "id": "85a7dadbec463d65" }, { "metadata": { "ExecuteTime": { - "end_time": "2025-01-10T13:57:54.328409Z", - "start_time": "2025-01-10T13:57:53.334623Z" + "end_time": "2025-01-14T16:28:13.419337Z", + "start_time": "2025-01-14T16:28:12.855842Z" } }, "cell_type": "code", @@ -838,9 +842,9 @@ "name": "stdout", "output_type": "stream", "text": [ - "Prediction: 1.8186865\n", + "Prediction: 1.7844329\n", "True value: 1.844\n", - "Average prediction: 2.0852828\n" + "Average prediction: 2.0879457\n" ] } ], @@ -850,32 +854,75 @@ "metadata": {}, "cell_type": "markdown", "source": [ - "### Traditional Explanation with Baseline Imputation\n", - "The traditional way to explain any black-box model trained on tabular data is by using imputation strategies for feature removal (excellent [paper by Covert et al.](https://jmlr.csail.mit.edu/papers/volume22/20-1316/20-1316.pdf)).\n", - "During explanations, the model is queried multiple times with different subsets of features removed.\n", - "Removed features are imputed using different strategies, such as the baseline imputation.\n", - "Baseline imputation replaces the removed features with the mean/mode of the training data.\n", + "### Explaining TabPFN with Remove-and-Contextualize\n", "\n", - "We can natively use the ``shapiq.Explainer`` (specifically ``shapiq.TabularExplainer``) to explain the TabPFN model:" + "Since TabPFN is a foundation model, it uses in-context learning to solve the classification and regression tasks.\n", + "This means that \"retraining\" the model is quite inexpensive, because we only need to provide the new data points, remove the features that are out-of-coalition, and re-contextualize the model with the new data points.\n", + "A recent paper by [Rundel et al.](https://arxiv.org/pdf/2403.10923) shows that this strategy is very effective for explaining models like TabPFN.\n", + "\n", + "This notion of remove-and-recontextualize is implemented in ``shapiq.TabPFNExplainer``:" ], - "id": "b225c897c1181eee" + "id": "cdba7867ce6fbbb0" }, { "metadata": { "ExecuteTime": { - "end_time": "2025-01-10T14:02:39.961668Z", - "start_time": "2025-01-10T13:57:54.329359Z" + "end_time": "2025-01-14T16:28:13.434649Z", + "start_time": "2025-01-14T16:28:13.421341Z" } }, "cell_type": "code", "source": [ - "explainer = shapiq.Explainer(model, data=x_test[:50], index=\"SV\", max_order=1, imputer=\"baseline\")\n", - "explainer._imputer.verbose = True # see the explanation progress\n", - "\n", - "shapley_values = explainer.explain(x_explain)\n", - "shapley_values.plot_force(feature_names=feature_names)" + "explainer = shapiq.Explainer(\n", + " model=model,\n", + " data=x_train,\n", + " labels=y_train,\n", + " index=\"SV\", # Shapley values\n", + " max_order=1, # first order Shapley values\n", + " empty_prediction=float(average_prediction), # Optional, can also be inferred from the model\n", + ")\n", + "print(f\"Explainer Class: {explainer.__class__.__name__} inferred from the model.\")" ], - "id": "41314e231db2e986", + "id": "a09303a6df37811", + "outputs": [ + { + "name": "stdout", + "output_type": "stream", + "text": [ + "Explainer Class: TabPFNExplainer inferred from the model.\n" + ] + } + ], + "execution_count": 9 + }, + { + "metadata": {}, + "cell_type": "markdown", + "source": [ + "When we initialize the ``shapiq.Explainer`` with the TabPFN model, ``shapiq`` automatically infers the explainer class from the model and returns the correct ``shapiq.TabPFNExplainer``.\n", + "The ``shapiq.TabPFNExplainer`` is more of a wrapper of the ``shapiq.TabularExplainer`` with the distinction that it uses the TabPFNImputer to apply the remove-and-recontextualize strategy for model explanation.\n", + "In the following, we will precompute the values of the imputer for our explanation data point such that we can quickly explain the TabPFN model with different explanation methods. Note that this is not necessarily needed and you can just call ``explainer.explain(x_explain)`` as we will do later." + ], + "id": "9dda61d4a5137375" + }, + { + "metadata": { + "ExecuteTime": { + "end_time": "2025-01-14T16:30:51.853358Z", + "start_time": "2025-01-14T16:28:13.436654Z" + } + }, + "cell_type": "code", + "source": [ + "imputer = explainer._imputer\n", + "if not os.path.exists(\"tabpfn_values_explainer.npz\"):\n", + " imputer.verbose = True # see the pre-computation progress\n", + " imputer.fit(x_explain)\n", + " imputer.precompute()\n", + " imputer.save_values(\"tabpfn_values_explainer.npz\")\n", + "imputer.load_values(\"tabpfn_values_explainer.npz\")" + ], + "id": "78f5c9bcdb9c44a8", "outputs": [ { "data": { @@ -885,80 +932,217 @@ "application/vnd.jupyter.widget-view+json": { "version_major": 2, "version_minor": 0, - "model_id": "58adb18d135f41429ff10942996b0c2a" + "model_id": "11d22bd5ec8b4d28ac2088a9a3bca064" } }, "metadata": {}, "output_type": "display_data" + } + ], + "execution_count": 10 + }, + { + "metadata": {}, + "cell_type": "markdown", + "source": "Let's now explain the TabPFN model with the Shapley values and visualize the results:", + "id": "8bc8070634c643c1" + }, + { + "metadata": { + "ExecuteTime": { + "end_time": "2025-01-14T16:30:52.390677Z", + "start_time": "2025-01-14T16:30:51.855390Z" + } + }, + "cell_type": "code", + "source": [ + "shapley_values = explainer.explain(x_explain)\n", + "display(shapley_values.dict_values)\n", + "shapley_values.plot_force(feature_names=feature_names)" + ], + "id": "5581b09a2b92d2fa", + "outputs": [ + { + "data": { + "text/plain": [ + "{(): 2.0879456996917725,\n", + " (0,): -0.16839496683757907,\n", + " (1,): 0.012176953230593706,\n", + " (2,): -0.06017028761354257,\n", + " (3,): 0.06772990684959702,\n", + " (4,): 0.011347168594688171,\n", + " (5,): 0.008271948812239768,\n", + " (6,): -0.10029863386281736,\n", + " (7,): -0.07417490089998008}" + ] + }, + "metadata": {}, + "output_type": "display_data" }, { "data": { "text/plain": [ - "
" + "
" ], - "image/png": "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" + "image/png": "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" }, "metadata": {}, "output_type": "display_data" } ], - "execution_count": 9 + "execution_count": 11 }, { "metadata": {}, "cell_type": "markdown", + "source": "We can also explain the TabPFN model with the Faithful Shapley Interaction values:", + "id": "cae4140040973f53" + }, + { + "metadata": { + "ExecuteTime": { + "end_time": "2025-01-14T16:30:52.945174Z", + "start_time": "2025-01-14T16:30:52.391669Z" + } + }, + "cell_type": "code", "source": [ - "### Explaining TabPFN with Remove-and-\"Retrain\"\n", - "\n", - "Since TabPFN is a foundation model, it uses in-context learning to solve the classification and regression tasks.\n", - "This means that \"retraining\" the model is quite inexpensive, because we only need to provide the new data points with whatever features we want to remove.\n", - "A recent paper by [Rundel et al.](https://arxiv.org/pdf/2403.10923) shows that this strategy is very effective for explaining models like TabPFN.\n", - "\n", - "Because of ``shapiq``'s notion of cooperative games, we can easily implement the remove-and-\"retrain\" strategy for TabPFN as game.\n", - "The game takes the model, the training data, the explanation data, and the average prediction as input.\n", - "The value function of the game performs the remove-and-\"retrain\" strategy for TabPFN and returns the predictions for the coalitions." + "explainer = shapiq.Explainer(\n", + " model=model,\n", + " data=x_train,\n", + " labels=y_train,\n", + " index=\"FSII\", # Shapley values\n", + " max_order=2, # first order Shapley values\n", + " empty_prediction=float(average_prediction), # Optional, can also be inferred from the model\n", + ")\n", + "# let's just make sure we use the precomputed values for speedup with the CPU\n", + "explainer._imputer.load_values(\"tabpfn_values_explainer.npz\") # Optional\n", + "fsii = explainer.explain(x_explain)\n", + "display(fsii.dict_values)\n", + "fsii.plot_force(feature_names=feature_names)" ], - "id": "cdba7867ce6fbbb0" + "id": "b807df57f73586af", + "outputs": [ + { + "name": "stderr", + "output_type": "stream", + "text": [ + "C:\\1_Workspaces\\1_Phd_Projects\\shapiq\\shapiq\\approximator\\regression\\_base.py:342: UserWarning: Linear regression equation is singular, a least squares solutions is used instead.\n", + "\n", + " warnings.warn(\n" + ] + }, + { + "data": { + "text/plain": [ + "{(): 2.0879456996917725,\n", + " (0,): -0.21091068074703873,\n", + " (1,): 0.007484612011888939,\n", + " (2,): -0.09822735161895646,\n", + " (3,): 0.12143328814392403,\n", + " (4,): -0.031076746894069238,\n", + " (5,): -0.011552190779500905,\n", + " (6,): -0.0713138114827143,\n", + " (7,): -0.02120219355424014,\n", + " (0, 1): -0.011577214797903055,\n", + " (0, 2): 0.11770563437769158,\n", + " (0, 3): -0.13197785644361207,\n", + " (0, 4): -0.008421346403875199,\n", + " (0, 5): -0.03473570715869672,\n", + " (0, 6): 0.07825948312112639,\n", + " (0, 7): 0.07577844489132672,\n", + " (1, 2): -0.007361223300846822,\n", + " (1, 3): -0.01226908337559184,\n", + " (1, 4): -0.021566055786592563,\n", + " (1, 5): 0.04048377247029388,\n", + " (1, 6): 0.022422441130022785,\n", + " (1, 7): -0.0007479466148791094,\n", + " (2, 3): 0.005479536453317312,\n", + " (2, 4): 0.013502439146366065,\n", + " (2, 5): -0.03144936760323395,\n", + " (2, 6): -0.010556209371776067,\n", + " (2, 7): -0.01120668678111711,\n", + " (3, 4): -0.0012490939531917633,\n", + " (3, 5): 0.04103636032002048,\n", + " (3, 6): 0.001323158115996492,\n", + " (3, 7): -0.009749784072804244,\n", + " (4, 5): 0.036554256222930834,\n", + " (4, 6): 0.035570151180879,\n", + " (4, 7): 0.030457457189895433,\n", + " (5, 6): -0.0033763905371929697,\n", + " (5, 7): -0.008864646582937265,\n", + " (6, 7): -0.18161225858244862}" + ] + }, + "metadata": {}, + "output_type": "display_data" + }, + { + "data": { + "text/plain": [ + "
" + ], + "image/png": "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" + }, + "metadata": {}, + "output_type": "display_data" + } + ], + "execution_count": 12 + }, + { + "metadata": {}, + "cell_type": "markdown", + "source": [ + "### The Remove-and-Recontextualize Strategy Explained\n", + "Because of ``shapiq``'s notion of cooperative games, we can easily implement the remove-and-recontextualize strategy for TabPFN as a cooperative game.\n", + "The game takes the model, the training data, the explanation data, and the empty prediction (average prediction) as input.\n", + "The value function of the game performs the remove-and-recontextualize strategy for TabPFN and returns the predictions for the coalitions." + ], + "id": "6c61da3b9399a6aa" }, { "metadata": { "ExecuteTime": { - "end_time": "2025-01-10T14:02:39.977683Z", - "start_time": "2025-01-10T14:02:39.963673Z" + "end_time": "2025-01-14T16:30:52.961182Z", + "start_time": "2025-01-14T16:30:52.947201Z" } }, "cell_type": "code", "source": [ "class TabPFNGame(shapiq.Game):\n", - " \"\"\"The TabPFN Game class implementation a remove-and-\"retrain\" strategy to explain the predictions of TabPFN.\n", + " \"\"\"The TabPFN Game class implementation a remove-and-contextualize strategy to explain the predictions of TabPFN.\n", + "\n", + " Note:\n", + " This is a simplified implementation of :class:`shapiq.TabPFNImputer`.\n", "\n", " Args:\n", " model: The TabPFN model.\n", " x_train: The training data.\n", " y_train: The training labels.\n", " x_explain: The data point to explain.\n", - " average_prediction: The average prediction of the model.\n", + " empty_prediction: The average prediction of the model.\n", " \"\"\"\n", "\n", - " def __init__(self, model, x_train, y_train, x_explain, average_prediction):\n", + " def __init__(self, model, x_train, y_train, x_explain, empty_prediction):\n", " self.model = model\n", " self.x_train = x_train\n", " self.y_train = y_train\n", " self.x_explain = x_explain\n", - " self.average_prediction = average_prediction\n", + " self.empty_prediction = empty_prediction\n", "\n", " print(\"Initializing TabPFN Game\")\n", " print(\"Train data shape: \", x_train.shape, y_train.shape)\n", " print(\"Explain data shape: \", x_explain.shape)\n", "\n", - " super().__init__(n_players=x_train.shape[1], normalization_value=self.average_prediction)\n", + " super().__init__(n_players=x_train.shape[1], normalization_value=self.empty_prediction)\n", "\n", " def value_function(self, coalitions: np.ndarray) -> np.ndarray:\n", " \"\"\"The value function performs the remove-and-\"retrain\" strategy for TabPFN.\"\"\"\n", " output = np.zeros(len(coalitions), dtype=float)\n", " for i, coalition in enumerate(coalitions):\n", " if sum(coalition) == 0:\n", - " output[i] = self.average_prediction\n", + " output[i] = self.empty_prediction\n", " continue\n", " x_train_coal = self.x_train[:, coalition]\n", " x_explain_coal = self.x_explain[coalition].reshape(1, -1)\n", @@ -969,13 +1153,13 @@ ], "id": "37a977c5f4a88aee", "outputs": [], - "execution_count": 10 + "execution_count": 13 }, { "metadata": {}, "cell_type": "markdown", "source": [ - "With this game implementation we can now use helper functions from ``shapiq.Game`` like ``precompute`` to precompute the values of the game to speed up the explanation process.\n", + "Similar to the above imputer, with this game implementation we can now use helper functions from ``shapiq.Game`` like ``precompute`` to precompute the values of the game to speed up the explanation process.\n", "For reproducibility, this notebook loads a precomputed game from the file ``tabpfn_values.npz`` if it exists." ], "id": "c8b473a6c67a54a2" @@ -983,25 +1167,47 @@ { "metadata": { "ExecuteTime": { - "end_time": "2025-01-10T14:02:39.993671Z", - "start_time": "2025-01-10T14:02:39.980669Z" + "end_time": "2025-01-14T16:33:29.874001Z", + "start_time": "2025-01-14T16:30:52.965176Z" } }, "cell_type": "code", "source": [ - "import os\n", - "\n", - "if not os.path.exists(\"tabpfn_values.npz\"):\n", + "if not os.path.exists(\"tabpfn_values_game.npz\"):\n", " tabpfn_game = TabPFNGame(model, x_train, y_train, x_explain, average_prediction)\n", " tabpfn_game.verbose = True # see the pre-computation progress\n", " tabpfn_game.precompute()\n", - " tabpfn_game.save_values(\"tabpfn_values.npz\")\n", + " tabpfn_game.save_values(\"tabpfn_values_game.npz\")\n", "\n", - "tabpfn_game = shapiq.Game(path_to_values=\"tabpfn_values.npz\", normalize=False)" + "tabpfn_game = shapiq.Game(path_to_values=\"tabpfn_values_game.npz\", normalize=False)" ], "id": "7b2606969b5bab0", - "outputs": [], - "execution_count": 11 + "outputs": [ + { + "name": "stdout", + "output_type": "stream", + "text": [ + "Initializing TabPFN Game\n", + "Train data shape: (50, 8) (50,)\n", + "Explain data shape: (8,)\n" + ] + }, + { + "data": { + "text/plain": [ + "Evaluating game: 0%| | 0/256 [00:00" + "
" ], - "image/png": "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" + "image/png": "iVBORw0KGgoAAAANSUhEUgAABdIAAAF1CAYAAADsoBjiAAAAOnRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjEwLjAsIGh0dHBzOi8vbWF0cGxvdGxpYi5vcmcvlHJYcgAAAAlwSFlzAAAPYQAAD2EBqD+naQAATZ5JREFUeJzt3QecnFW5P/Bn+6YXEgIESEjovRcBAUFARLGC2LuiXvUievXa8Hq92Av+r+3aGypYsKEC0qv03pNQQwmkb9/5f867mc3OJhnYzc7s7Mz362fd7OzsvO9wzjvnnd+c9zl1uVwuFwAAAAAAwHrVr/9mAAAAAAAgEaQDAAAAAEARgnQAAAAAAChCkA4AAAAAAEUI0gEAAAAAoAhBOgAAAAAAFCFIBwAAAACAIgTpAAAAAABQhCAdAAAAAACKEKQDAAAAAEARgnQAAAAAAChCkA4AAAAAAEUI0gEAAAAAoAhBOjB6fnFxxOTXRFx2+7Pfd9f3RRz3meFv63/O7tvWoieG/xj0u+rBjpjzhUfj7FtXj/auAAAAAJScIB2gGv35XxEr2kZ7LwAAAACqQuNo7wDAc3L91yLq6kZ7L8aGR5+OeMuZEXtuE/G7j0VMGjfaewQAAAAwppmRDowNLU0RzWPos7/RnA2+xfSIn3wg4sYHIl75+YiV7aO3LxWmpzcXbV29o70bAAAAwBgzhlIpoGr15iLO/FPE98/vm0291YyI014e8brDCmukbz0z4q+fLvzb7/8j4lvnRTz4ZMSWMyJOOTZiYmvEKd+J+MsnIw7dpfD+nd0Rnzkr4qzLIp5aHrH9FhGfPjnimL3W3a/fXhnx3b9H3LYoJbARO28V8YGXRLzswML7pdrrr31+xGsOjfifcyJuXRix17x193WkpHrvz8Xe8yOuvjvilWdE/OHjEeOao5RWd/bGN69aGX++qy0Wr+iJKa31cejclvjQoZNiyyl9w01Hdy52+8ZjcfyO4+KrL57W/7cf+9vS+OXNq+Mt+0yI04+a0n/7e899Oi5+oCNu/sBm0Vjfd0XC8o7e+N+rVsZ5d7fFYyt6YmJzfRwytyU+/PxJsfXUtcNaqt9+2l+Xxi9O2iSuf6QzzrltdTy6vCc+f+zUePVu40v63wIAAACoLoJ0YPR95lcR7Z0Rbz2qb9b5Dy6IOOXbEfM3izhwhw3/3dfOjfj0WX0lTE4/OWJ1R8SZf46YMXnDf/Oub0U0NUT82/ERXd19Ifxrvxxxw9ci5my69n7/9euIL/8+4qg9Ij5+YkQKcVPd8Td+PeLLb4l45zGFj5tmf//x2og3vaAvVC+lz/92aPf/130RTy7r+yCiRLp6cvGG3zwd1z3SGcft0Brv2G9iLHymO35+46q4bGFH/OmNM2PzyQ3R0lgX+8xujqsWdRb8/RWLOrL/xFcu6ui/LZfLxdUPdsZ+WzYXhOiv+NlT8eiKnjhxt/Gx/YzGeGJlb/zsxlVxwk874k9vmtEf2ud97qJl0dUTcfIe47PQfd50Qx8AAAAwNNIEYPSlWeIX/8/a0i1pxvfu7++bDb6hIP3plRFnnBOxy9YR//hMROua2dYpyN7n3ze8rU0mRfzmI2vrracZ60d8POJHF/aF8clNC/pC9A+d0DdbPe+UF0Wc/OW+4P/k5xfWHr/z4YhzPx5xxG5Rcst/9ez3SR8qnPTFiCvuivjhv5U0RE/OuXV1FqK/a/8J8Z9HrJ1RfvDclnjrOU/HFy5dHl8/vm8G+vO2bokrF62IBU93xzbTG+OR5d2xaGlPvHyXcfH729viyVU9MXNCQ9z9VHc8tbo3njenpf/xvnrZinhwWXf84Q0zY+dNm/pvf9Vu4+KYHz4ZX7t8RXxlwEz3pL07F39988wY16SaGQAAADA8UgVg9L39hYX1z1ON7203j3hg8Yb/5qJbItq7It521NoQPZk1NeLVh2z471IYPnDR0n3m95WCuf+xtbf95vK++7z2sIglywu/jtunr/75tfcUPu5uc8oToj8Xg0P0waVoSuBv97ZnM8rfe9CkgtuPnN8aO2/aGOff2x69uVx2Wz4Yv/LBvtnnVy7qjIa6iH8/eFKklsnPSr9qzff8/dMM9T/c0RYHbNkSm02sj6dX9/R/jW+qi722aI5LF66d0Z73+j0nCNEBAACAjWJGOjD65g4oqZI3fWLEQ09t+G8WPdn3fbst1v3ddpsPfVtphnve3Y+k1DZin1M3/DhPLCv8OQX/laKjq2+B0TKF6MlDS3ti1sT6rC76YNvPaIo7nuiOp1f3xowJDbHH5k0xsbkuC8xft+eE7PvumzXFnGmNsePMxixYP2Hn8XHlg50xtbUudtm0b6hasro3nmnrzcLyvb75+Hr3Y00FmAJp1jsAAADAxpAuAKOvYQOzhdfMYC77ttK/04z03350w/ffacvCn0u8kOf6tHX1xj/v74gX7zigxEwybWLEhZ+NqK/MWdip3nmqe37Vg53ZLPMUpL9y1/H9s8//sWb2+tUPdmQ/1625giDfQofMaY5TDiyc+V7MuKb1pOsAAAAAQyBIB8amfM3vex+NOGzXwt/dO6BMy3DM3zzigpsjtpoRscPsqETtXbl462+fjmse7IwdZzbF/E0GvZyXOUTfempDXLKgO5a1964zK/3eJd0xqbkupo9fe3sKyC96oCP+end7LF65tg76wXNa4gfXrYrz7m6P5R25gvrom4yvj8ktdbGyMxeHzF17OwAAAECpVeZ0RYBnk+qRtzRF/OCCiPbOtbc/vjTi7Ms37rFfc2jf97SoaE/vur9/YmmMfoi+JAvRv/LiqeuG6KPgmO1aozcX8e2rVxaWsr+/PW5/vCuO2q416gfUps8H5F+9fEW0NETsu2XfjP79t2rO6qWnRUOz+229dqZ/+vuX7TIubnqsK/5yV9t69+OpVT0leX4AAABAbRv99AVgODaZFPHRV/aF3Ud/OuLEQyLaOiN+fGHfjPIbHyhcVHQo0gKkH3tVxBnnRBz8H311xjefFrH4mYibFkT848aIJb+I0dDZk4u3/XZJXLGoM3bbrCkWPtMdX7t8+bP+XVoEtDkl1CXyqt3Gxzm3tcW3r1kZDy/rjv23asn27ec3roqZE+rjI8+fXHD/VPc81T+/b0l3HLh1c7Q29u3bpJb6rF76jY91xaYT62O7GU0Ff/fh50+O6x7ujPee+0ycd3dbtsBoU0NdPLK8Jwvt03+Tr7x4WsmeJwAAAFCbBOnA2PWhl0VMGhfx7fMiTj8rYssZEe8/vq+YdgrSWzeibnkK0veaF/Gdv0V8668RqzsiZk6O2GmriC++OUbL8vbe+NfDfTPwb13clX09F+/cf2JJg/QUZv/sxOnxzatWxp/ubIu/3dMek1vr47gdx8Vph06KLSY3FNw/1T0/cOuW7H7P27qwTMvz5rZkQfpBg25PJrfUx+9ePyO+d+2qbFb6P+5rz2qubzaxIau7/po9+mqtAwAAAIykulxa6Q2gmpz2o4jv/T3i3u9EzJoa1ebC+9vjlN8/HTtv2hQ/PWmTLFwGAAAAoHSkL8DYNbA2el4qv/KrSyN23qoqQ/TkyPmt8e2XT487nuiKN/56SSzvWE8ddwAAAABGjNIuwNh12R0Rn/xFxEv2j5g9PeLBJyN+/M+Ile0Rp58c1Swfpn/50uXR0Z2LWLcKCgAAAAAjRGkXYOy6f3HEJ34ecf39EU+viGht6qtrfurLIo7YLWpBby4X9cNdVBUAAACA50SQDgAAAAAARaiRDgAAAAAARQjSAQAAAACgCEE6AAAAAAAUIUgHAAAAAIAiBOkAAAAAAFCEIB0AAAAAAIoQpAMAAAAAQBGCdAAAAAAAKEKQDgAAAAAARQjSAQAAAACgCEE6AAAAAAAUIUgHAAAAAIAiBOkAAAAAAFCEIB0AAAAAAIoQpAMAAAAAQBGCdAAAAAAAKEKQDgAAAAAARQjSAQAAAACgCEE6AAAAAAAUIUgHAAAAAIAiBOkAZXT22WfHu9/97th3332jpaUl6urq+r+G6vDDDy/4+2Jfg91+++3xzne+M3bccceYMGFCNDY2xtSpU7P9+sQnPhFPPPHECD1jAAAYeaeffnp2nvvUU09Frf83AKA8Gsu0HQAi4nOf+1zcfPPNo7oPf/3rX+PlL395dHZ2Fty+bNmyuP7667OvH/zgB3HNNdfE1ltvPWr7CQAAAFApBOkAZZRmjMyfPz+b+b148eK45JJLhv1Yp5xyShx//PHr3N7d3R2f/OQns+/Ji170ooLff/jDH+4P0dOs+Le+9a0xe/bsuPDCC+Oiiy7Kbk/79v/+3/+LL37xi8PePwAAAIBqIUgHKKMrr7wyxo0b138p5sYE6SeddNJ6b//lL3/ZH6InH/nIRwp+f//99/f/+x3veEd885vfzP790Y9+NDbZZJNsZnry5JNPDnvfAAAAAKqJGukAZZQP0Uvpy1/+cv+/08z3VEt9oJ133rn/31dccUU88MAD0dHRkZV8WbFiRf/vjj322JLvKwAAbIxUI/3EE0+MyZMnZ5NCPvCBD0R7e3vBfX70ox/FC17wgth0002zKzLT+fC3v/3tdR7ruuuui2OOOSZmzJiRnbdvs8022dWbA/X29sbXv/712GWXXaK1tTVmzZoV73rXu+KZZ5551nP0dHXqokWL1vndxz72sWhubu5/jMsuuyxe/epXZ2UW0/5utdVW8e///u/R1tZWdBsLFy7MtvHjH/94nd+l29NEnoEeeeSR7Pml55C2k57TD3/4w6LbAKhlZqQDVJELLrggbrzxxg3ORk/SiX8qCZNC83TfVGpmoOnTp2cLjm5oxjsAAFSKFKLPnTs3zjjjjLj66qvjzDPPzALpn/70p/33SaF5Colf+tKXRmNjY/zpT3+K97znPVko/t73vje7zxNPPBFHH310zJw5M7tSc+rUqVkw/bvf/a5geyk0T0H1W97ylnj/+98fCxYsyEoipvPqNEmlqalpg/uZzs1/85vfZKUWB0q3pW1PmzYt+/nss8+O1atXZ6Uc04cD1157bXYV6cMPP5z9biQ8/vjjceCBB2YB+/ve977seZ933nnxtre9LZYvXx4f/OAHR2Q7ANVEkA5QRb70pS/1/3vevHnxile8Yp37PP/5z88WEn3Zy14W99xzzzq/TyF7+h0AAFS6NGv83HPPzf6dQvE0M/1b3/pWnHbaabH77rtnt6dyigOvDE3Bcbr68qtf/Wp/kJ5KMKYA/h//+Ed2VWfef//3f/f/+/LLL4/vf//78Ytf/CJe+9rX9t9+xBFHZI+XQu6Btw+UZpen4PrXv/51QZD+r3/9K7tCdOBs8S984QsF+/vOd74ztt122/jP//zPePDBB7PH2lgf//jHo6enJ2699dYsrE/e/e53x8knn5ztS/rAoBxX0wKMJUq7AFSJW265JTvxzzv11FOjoaFhnfulNxKHHHJIFqKnGTlvfvOb4zOf+UwcdNBB2e/T7J399ttvvSE7AABUknwQnvdv//Zv2fdUtjBvYCCc1gNK5WAOO+ywLMDOrw+UZqAnf/7zn6Orq2u920pB+ZQpU+KFL3xh9hj5r3322ScmTpwYF110UdF9TVd8Xn/99QVrFqVgPZVVOeGEE9a7v6tWrcq28bznPS9yuVzB1afDlR7nt7/9bbzkJS/J/j3wuaTSNum/yQ033LDR2wGoNoJ0gCoxsDZ6qus4uJ5jkmqhv+51r4unn346+/mTn/xkVjPyU5/6VFx66aWx3XbbZbcvWbIk/uu//quMew8AAEOXP3/NS2UL6+vrs7IseankylFHHRUTJkzIAvNUxiTN7k7yQXoK1l/5yldmE0zSuXQKttN5cjp/zrv33nuz+6da6+kxBn6tXLkyKw9TTKp7nvYthedJCrFTOP+iF70om0mfl2adp8kuqeRiCujT46f9G7i/G+PJJ5+MpUuXxve+9711nkcqWZM823MBqEVKuwBUgVQv8Ve/+lXBzJz1XYp51113ZYsK5aWZ53lpdvqee+6ZvUFIbrrpppLvNwAAjKRU83ugNPv7yCOPjB133DEr5ZIW7kwLe6YZ61/72teyOun5vzvnnHOyOuuphvrf//73bGLKV77yley2FGin+6YQPZV2WZ8URBezxRZbxKGHHprVRE9BfnrcFJqnUi55qdxKmvGeJr78x3/8R7bf6QOAdA6fwvX8/j6X5z7wMQfKP8brX//6eNOb3rTev8mXxQFgLUE6QAXLL2SUl2atrE9aQDR/CWoK0FPdx+dyEp1qMqYZMEl3d3dBeK4mIgAAlS5NAkl10vPuu+++LChOC5AmKRRPs8r/+Mc/FtQW31AZllTHPH197nOfi1/+8pfZ1Zxpwsrb3/72bLb7BRdcEAcffPCwz5VTeZe00Ondd9+dzUwfP358VmIlL9UsTyUWf/KTn8Qb3/jG/tvPP//8Z33s/GKlabb5QIsWLVon8J80aVL23iDN1AfguVHaBaCMvv3tb2cLH6WvgfXMk/zt6Wtg3cRns3z58vi///u//p9T8J4uR12fXXfdtX8xoeSzn/1sNtMmlXFJi5DmZ6MnqT4iAABUsv/93/8t+Pmb3/xm9j0/WSS/ZtDACSmpPEoq2zJQWmh08KSVdLVmki/vcuKJJ2bhczqHHixNShkcYK9PKh+T9umss87Kyrocf/zx2YzzvPXtb/r3N77xjWd97FQeJr0PSCUbB0qLrw6UtpH2I9VJv+2229Zb+gWAdZmRDlBGadZJWuxzfdJlo3nphDrNeHkuvvvd72Zhev6k+EMf+tAG75suY/3Od74TJ598cnayn74Gv4lI9thjj/jIRz7ynLYPAACjZcGCBfHSl740jj322Ljqqqvi5z//ebz2ta/NzmeTo48+OjsHTrO+3/Wud2W1zNMklFSi5bHHHut/nDQDPAXOL3/5y7Pz8BUrVmT3S+H0cccdl90n1SlPj3HGGWdkV3Kmx25qasomo6RQPIXdr3rVq4rub9ruEUcckZWZSdtIM9QHSqVc0vbT5JpUziVtPwXeKeh/LtLM+c9//vPZ93333TcL1dMM98HSfdKs/AMOOCDe8Y53xM4775yVk0mLjKZZ9/k1lQBYS5AOMIalci4DZ6e84hWviHnz5hX9m3Ryv/3222ezddKJdaqvnmbZpJP0dAKd3jykGuutra1leAYAALBxE1U+9alPxUc/+tFszZ9U4vBLX/pS/+932GGHrPb5Jz7xiSyc3myzzeKUU07JypukKzPzUkh+7bXXZmVcHn/88ZgyZUrsv//+WT30gaVj0qSUffbZJ5vMkuqcp22mMjKp3ngq+fJcpPA8hdWpvEo+pM9LwXwqR/P+978/C+zTOXk6P0/PK//hQDHpv0WaUZ6ec6rFnmbmn3feeVmAP9CsWbOy55uuTP3d736XfYiQrlzdZZddCmq2A7BWXW5DBXcBAAAAAAA10gEAAAAAoBhBOgAAAAAAFCFIBwAAAACAIgTpAAAAAABQhCAdAAAAAACKEKQDAAAAAEARgnQAAAAAAChCkA4AAAAAAEUI0gEAAAAAoAhBOgAAAAAAFNFY7JcAVL7Vq1dHLpeLurq6GD9+/GjvDgAAMIDzdYDqIEgHGON6e3v7T8wBAIDK4nwdoDoo7QIAAAAAAEUI0gEAAAAAoAhBOgAAAAAAFCFIBwAAAACASgjSL7300njJS14SW2yxRbbAxh/+8Idn/ZuOjo74+Mc/HnPmzImWlpaYO3du/PCHP+z//Y9//OPssQZ+tba2lviZMNJt/eY3v3mddkxfu+yyS8H9/vd//zfrA6mNDzjggLj22mtL/EwYjbY+/fTT1/n9jjvuWIZnw0i/hv/iF7+IPfbYI8aPHx+bb755vPWtb40lS5YU3Ofss8/O2jcd17vttlv89a9/LeGzYDTb25hdPW2dxuOddtopxo0bFzvssEP89Kc/Xec+ju3aaGvHdWU644wzYr/99otJkybFpptuGi972cvi7rvvfta/e7bjNi2U+KlPfSp7jU994qijjop77723hM+E0Wrr9Z2zH3vssSV8JpSirW+//fZ45Stfmb2HTm349a9/fb338z67dtrbe+3qaOv/+7//i0MPPTSmTZuWfaXxePBxa8yugiB91apV2Rvs9CL9XJ144olx4YUXxg9+8IOsI5111lnZSfxAkydPjscee6z/a9GiRSXYe0rZ1t/4xjcK2vChhx6K6dOnx6tf/er++/z617+OU089NT796U/HDTfckD3+McccE0888UQJnwmj0dZJCtYH3u/yyy8v0TOgVG19xRVXxBvf+MZ429velp3UpTdsaXB/xzve0X+fK6+8Mk4++eTsPjfeeGN20pC+brvtthI+E0arvRNj9thv629/+9vxsY99LHsjltr6M5/5TLz3ve+NP/3pT/33cWzXTlsnjuvKc8kll2RtdfXVV8f5558fXV1dcfTRR2d9YEOey3H7xS9+Mc4888z4zne+E9dcc01MmDAhOx9vb28v0zOjXG2dpOB84LGd3oszttp69erVMW/evPj85z8fm2222Xrv4312bbV34r322G/riy++OHsdv+iii+Kqq66KrbbaKvubRx55pP8+xuwSyo2CtNnf//73Re9z3nnn5aZMmZJbsmTJBu/zox/9KLsPleu5tPVg6f51dXW5hQsX9t+2//7759773vf2/9zT05PbYostcmecccaI7i+j39af/vSnc3vssUcJ9rB6rVixIrd8+fLse6W09Ze+9KXcvHnzCm4788wzc7Nnz+7/+cQTT8y9+MUvLrjPAQcckHvXu941wntMJbS3Mbs62vqggw7KnXbaaQW3nXrqqbmDDz64/2fHdu20teN6bHjiiSeyNr/kkks2eJ9nO257e3tzm222WfZ6n7d06dJcS0tL7qyzzirh3lPutk7e9KY35U444YSS7mutGenz9efS1gPNmTMn97WvfW2d273PHhtGqr29166+tk66u7tzkyZNyv3kJz/JfjZml1bF1kj/4x//GPvuu2/2Kcrs2bNj++23j9NOOy3a2toK7rdy5cqs9Ev6BOaEE07IZswwtqUrENJlJ6ldk87Ozrj++uuz2/Lq6+uzn9Onb1RPW+elS47SpefpE/XXve518eCDD47aPjI8Bx10UHbFQbpUOGU2jz/+eJxzzjlx3HHH9d8nHb8Dj+skfUruuK7O9k6M2WNfKrs3uHRHulw0XYGQZtAkju3aaevEcV35li1bln1PVwFuyLMdtwsWLIjFixcX3GfKlClZGQjHdnW19cAZj6nMQLoi/JRTTlmnPB+V39bPxvvs2mrvPO+1K9tw2jpdjZDOzfJ/Y8wurYoN0h944IHsEpN0idnvf//7rL5TelP+nve8p/8+aVBPNdPPPffc+PnPfx69vb3xvOc9Lx5++OFR3XeG79FHH43zzjsv3v72t/ff9tRTT0VPT0/MmjWr4L7p5/TiQPW0dZJe3FPN1b/97W/ZZeVpEEj1v1asWDFq+8rQHXzwwVnN7JNOOimam5uzywvT4D2wpEA6fh3XtdPexuzqkMKW73//+9kb7/ShyXXXXZf9nE7e03idOLZrp60d15UvtckHP/jB7HV611133eD9nu24zX93bFd/W+fLuqQ1EVKZ1S984QtZ6YEXvehF2Xsyxk5bPxvvs2urvRPvtauzrf/jP/4j+3AkH5wbs0urMSq4A6WFD9Ib8/RmPPnqV78ar3rVq+Jb3/pWNiMmzYBLX3npxD0tiPTd7343PvvZz47i3jNcP/nJT2Lq1KlZnT5qs63TSXre7rvvng32aabbb37zm6yWI2PDHXfcER/4wAeyBU5SGJPq7334wx+Od7/73dmVCNReexuzq8MnP/nJ7AT8wAMPzMLVdEL+pje9KbuCMM1io7ba2nFd+VLd1TQxSQ3c6jeSbf2a17ym/99pMdJ0Tj5//vxslvqRRx650Y/PxnFc15aRbG/vtauvrVNN/F/96lfZ67MF38ujYt/xpJVlU0mXfIiepBPzdCK/oVkuTU1Nsddee8V9991Xxj1lpKS2TbOa3vCGN2QzGvNmzJgRDQ0NWamAgdLPxRbRYOy19fqksD2VdnJcj73Vx9Mn6SlMTSdpKVxNH4Kmdk8ha5KOX8d17bT3YMbssSlNZEjtmi4hXbhwYXY58Ny5c2PSpEkxc+bM7D6O7dpp68Ec15Xlfe97X/z5z3/OFiPbcssti9732Y7b/HfHdvW39fqkEhDpPZlje2y19bPxPru22nt9vNce22395S9/OQvS//GPf2TvwfKM2TUapKc35Kn0Q6q7mHfPPfdkM2A21KnSZUm33nprFsIz9qRLBtML+OBPQlPQus8++2SXFg68YiH9PHAWFGO/rdcnvQbcf//9jusxJgUvg2enphP1pG+du76ZjAOP6yStVO64rs72HsyYPbalwDSdj6V2TrNgjj/++IJZyo7t2mjrwRzXlSG97qY35Kk85j//+c/YZpttnvVvnu24TY+R3nwPvM/y5cvjmmuucWxXWVuvT5rIlmqkO7bHVls/G++za6u918d77bHb1ukKwXT1XyrTk9aXHMiYXWK5MkmrU994443ZV9rsV7/61ezfixYtyn7/0Y9+NPeGN7yh4P5bbrll7lWvelXu9ttvz1as3W677XJvf/vb++/zmc98Jvf3v/89d//99+euv/763Gte85pca2trdn9Gz1DbOu/1r399tmL8+vzqV7/KVhj+8Y9/nLvjjjty73znO3NTp07NLV68uOTPh/K29Yc+9KHcxRdfnFuwYEHuiiuuyB111FG5GTNmZKtXs+F2WL58efa9Utr6Rz/6Ua6xsTH3rW99K3uNvvzyy3P77rtvbv/99++/T2rfdJ8vf/nLuTvvvDNbRb6pqSl36623lux5MHrtbcyujra+++67cz/72c9y99xzT+6aa67JnXTSSbnp06dnr9l5ju3aaWvHdWU65ZRTclOmTMnOpx577LH+r9WrV/ffJ7V1avOhHLef//zns/Pvc889N3fLLbfkTjjhhNw222yTa2trK/tzpHRtnV4rTjvttNxVV12VHe8XXHBBbu+9987ei7e3t4/K86wGG3u+Ppy27ujo6H/d33zzzbN2Tf++9957++/jfXZlKlV7e69dHW2dxuPm5ubcOeecU/A3A19fjNmlU7Yg/aKLLspO2gd/velNb8p+n74fdthhBX+TBvZ0YI8bNy4L1U899dSCzvTBD34wt/XWW2cdaNasWbnjjjsud8MNN5TrKTGCbb106dKsnb/3ve9t8HG/+c1v9rd3Cmeuvvrqkj8Xyt/W6Y16GvhTO8+ePTv7+b777ivL8xmryhGkD6etzzzzzNzOO++ctXdq09e97nW5hx9+uOA+v/nNb3Lbb7991t677LJL7i9/+UvJngOj297G7Mo01LZOb7L33HPPrJ0nT56cnZTfdddd6zyuY7s22tpxXZnW187pK33omZfaOt/2z/W47e3tzX3yk5/M2joFb0ceeWT2gQvV1dbp/fbRRx+dmzlzZhawz5kzJ/eOd7xDsDrK5+vDaesUlq7vbwafw3mfXXlK1d7ea1dHW6fX5fX9TfpgNM+YXTp16f9KPesdgNJJl+Sll/K0QPPEiRNHe3cAAIABnK8DVIeKrZEOAAAAAACVQJAOAAAAAABFCNIBAAAAAKAIQToAAAAAABQhSAcAAAAAgCIE6QAAAAAAUIQgHQAAAAAAxmqQ3tHREaeffnr2neqmrWuHtq4d2rq2aO/aoa1rh7auHdq6tmjv2qGta4e2rh3aenTV5XK5XFSo5cuXx5QpU2LZsmUxefLk0d4dSkhb1w5tPfJWrlwZ6aW8rq4uJk6cGJVCW9cW7V07tHXt0Na1Q1vXFu1dO+fr2rp2aOvaoa1HV0XPSAcAAAAAgNEmSAcAAAAAgCIa4zlIlyCtWLEiRuNyhYHfqV7aunZo69JeKtrb2xuVQlvXFu1dO7R17dDWtUNb1xbtXTvn69q6dmjr2qGtS2fSpEnZ6/RG10jP198BAAAAAIBq8lzqzj+nIH20ZqQDMHYXGwUAAJyvA1TLjPTnVNolPYiVYAEqU319vRNzAACoUM7XAaqDxUYBAAAAAKAIQToAAAAAABQhSAcAAAAAgCIE6QAAAAAAUIQgHQAAAAAAimgs9ksAAAAAhq+urq7gOwBjkyAdAAAAoEQmTJgw2rsAwAhQ2gUAAAAAAIoQpAMAAAAAQBGCdAAAAAAAKEKQDgAAAAAARQjSAQAAAACgCEE6AAAAAAAUIUgHAAAAAIAiBOkAAAAAAFCEIL1WdXZHPLNy5B6vNxexZEVET+/IPeazbm95+bY32tsFAKgAvblcLGvrzb4D1WVZe2/0pPc7UEVyxi1G2IqO3ujq0Z9qVeNo7wBl1Nsb8fTKiAcWR7R3RnT1RHz13IjunuE/5qypEbvPjdhjbsS4lohzr4m4ZWGUzGbTIvbYJmK3ORHjmiN+e2XEHQ+Vbnv9203Pc5u+55q2+/urIm57sPTbhefiu++K2GRSLFndG28+96nR3hsAqtgOMxrjpN0nRHtXLm59vDNufqwrHl2xEeeSQEWY3FIXH3jepGy+0P1Pd8f1j3Rm30VFjHXbbtIYr91jQrR35+K2xZ1x8+KueGS5cYvhaaqPOPWQydn3xoa62HJKQ0xurYv6urrR3jXKRJBeS1JwftfDff+ur49oqY9oqI+47I7hP+bpr4nYf/u1P8/dNOK7f4+S+Z83ROy5zdqft54Z8YMLouQ++7qIfeYP2O6mEf93fum3C8/1CpP0rScX/3q4c7T3BoAq9sJtW7PvrU11sd+WLTG+qT4++velo71bwEY6cbfxUVdXF40NETvMbIrZUxrijb9ZEl0uxGWMO2JeS/a9tbEu9t2yJSa11Mdp5xm3GJ7Dt2mJlsa+0DxdwPPg0p7YadPGqG8Y7T2jXJR2qSUtTRHTJhTe9sI9N+4xz7+58Oc0W3zTKVEyF9xU+PPe8yOmTyzd9ja03b3mZTOAAQBqacbqfls2F9x24f3to7Y/wMh5wfy+sDHvsgUdQnTGvAnNdXHgVoV927jFxjhyzYSCvEktddHUYDZ6LRGk15pNp64bfM+YPPzHu+aeiBVthbcduXuUzBV3RazuWPtzmlF/xG5RclfdHbFqwIBbXxfxghI+TwCACnPYNq3RmM6B1ujozsXliwaclwFj0o4zG2P25MKL1YWNVIND57YUhJyprvVlC41bDM/MCfWx+2ZNBbdNGydWrTVavNZMn9QXPo9UIJzqq19yW+FtR+4RJdPRFXHFnYW3HVXC7Q0snTG4BE45tgsAUCGOHDRj9aoHO6KtSwVlGOuOnF84w/KR5d1x91N9pQOhmvr2NQ91xspO4xbDc8S81oJa6ClOS/XRqS2C9FqTQvSZg2agH7WRM6svGFTeZYvpETtvFSUzuJzMnE0jtts8Sm7w89xqRsQOs0u/XQCAUTZnakPM36RwFpYZqzD2NTf0zdod6ML7zNhl7Js9uSF2nGnconQlsKaOq7fIaA0SpNeiweVdttzIQPieRyMefLJ8s7VvXxSx+JnyzYLPu+OhiEeWDNqu8i4AQO3N6ntqVU/csrhr1PYHGBkHbNUSE5rXxgK9uVxc9ICwkbHvBYPGrWfaeuOGRztHbX8Y23aYsW4JrGnjhOi1SJBeiyaNi2gtXCgqXriRQfSFtxT+/PxdIloKX2RGTG492zt818iWmC+1fw7a7mG7RjRZnhkAqF7p0uXD5xUGEhc90BG9ro6HqvuQ7ObHuuKp1VYZZWzLKtjOK5w9fPED7cYthu2oQYuMpqt5xjcJ0muRIL0WpUtPZk0pvO3QXTYuEE4Bc8+AE67xLREH7Rglc+GgMiuTx0fsv13ptte/3VvW/VDigO1Lv10AgFGy9xbN6yym9U+Xx8OYN31cfey5eWHpC8c21SAtCDljQmG+oW8zXCk0P2RQCazp4+ujTlmXmiRIr1Uzp64bCB+4w/Afb8mKiJsWlK/cyuKlEbctKt/28p5YFnHzoOdp0VEAoIYWGb3rya54eHnPqO0PMDIOn9cSDWnq7hqru3rjqofUR6f6yrrcv6QrFi41bjE8+2/VEhMHlMDK10enNmn5WtXaFDFl/MgGwoNnie81L2KTSVG2RUf323bd51SORUf32TZi2oTSbxcAoMwmNtdlNZQHMqsPqrOsy+ULO6Kje9R2B0bEuKa6eN7WgxbQvd8HRAzfUYNeK9O5UXOD2ei1SpBey2YNmpW+9/yIaROH/3hX3RWxqn1QYbISLsZ5+R0R7QMWuUo10g/fLUruijsj2gYsUtJQX57tAgCU2aFzW6JpwJvFrp5cXLpQIAFj3XabNMbWUwvXtPqnsJEqcMiclmhpXDtudffm4pIFPgBm5EpgDS53R23R+rVsk8l9IXBe+vcLNiIQTtMXLrujfGVPUph95Z3l215eCu9TiD+Si7UCAIyBGavXPNQRqzqt1gbVdmw/tqIn7nhiwCQlqJK+fd3DnbG8w7jFyJTASv+c0mo2ei0TpNeyFJwPLr2ysXXGB5c92WpGxA6zN+4xh7K9+ZtFbDMryr7o6NxZfdsGAKgSW05uiB1mFs7Ccnk8jH2N9RHP32bdkk2iRsa6zSbWxy6zBo9bZqMzch/MpBC9fkCwTu0RpNe6weVd5m4ase3mw3+8Ox6KePTpwtuOLGF5l1sWRjy5rPyz0m9dGPH40vJvFwBglBZre6atN254dEB5O2BM2n/L5pjUUhgFXPSAsJHqG7eWt/fGdY8YtxiebddTAktZF/SAWjd5fERLU2kXHT1s14imhiiJ3ty6s8OP2K2wZE0ppOkag7eb6qSn6R0AAGNcttTN/MIZqxc/0J6degHVNcPy1sWd8fjK3lHbHxgJdesJ0lNt9G5dmxF6rWxuiJjQbDZ6rZP61bq6uohNp6wbfG9MIDw4YJ40LuKA7Yf/eEPd3tQJEftsW7rtbegDgynjI/bbrvTbBQAosT02a4pNxhdOhHB5PIx9U1vrYp/ZzQW3ObapBqmky6yJg8ct5cgYuRJYU8fVR13K0KhpgnQiNp06soHwE8sibl5QvrInjyzpKylTsL0SlpPJe+yZiNseHNka8wAAFWDwrL77lnTFoqU9o7Y/wMg4bJvWgoXz2rtyceUipS+ovtnDC5/pjvuf7h61/WFs22/L5pg8qASWsi4kegER45r7SrwMdNSeI7sIaJohPm1ClG12+AE79M2EL7XB291/u3X/WwIAjCHjm+rioK0HL0RoVh9UY9h45YMd0datZhNjW2tjxMFz1l1AF0bqtTKVdGlpNBsdQTp5g8u77Ldt38z04brizoi2ATMbUs3yVEO8VC69PaJzwKfNqSb7YbtEyV12R5rGsfbnxoaIw3ct/XYBAErkkDktBW8Wu3tzWZ1ZYGzbZlpDbDO9cOE8ZV2oBunD33FNa8etnt5cXLzAB8AMz5T1lMAyG508PYE+udy6P/dsxKoc9fV9q1QN1F3Cy4FTUD+4VlVXGS4/Ts+xoYzPEwCgxLoGrSiaznQGloIAxqb1vb3bmKWxoFIMXlC0bj1v0+G5SqdB60Zkrtyhj2GTtXXNB7r6noiVGzE74dCdI1qaCkPtS26LkkkLpKZZ6HkdXX2zxUst2+6AWR1pVnyaHQ8AMEZdlUo9dK19w5hC9MMHLbgFjD0PLuuJe5/qKlq+AMaiax7qiJWda9P0+rq6OGKevs3wrOjIxbUPF64d8UybIJ0+gnQiVndErGgrXuN8qAYvLnrtPRHLB21jJA3e3lV39T2vUhu83avv3rgPIAAARll7d8QVizqKLj4KjE0XDCrlcuDWLTFhQEkMGIs6eyIuXzh43PIBMMN34X2Fr5Wru3LRbj0JBOlknlha+PPTKyOuv2/4j7fF9Ihdti6+KOdImjMzYrstCm+74JYoua1mROwwe9B2S/g8AQDKZHDd5LnTGmP+oNrKwNhz2cKO6OpZGwY1N9TFoXMFjlTfuLXllMbYYYZxi+G54dHOeKatsGbQM6s3ovwxVUOQXutSnafBZV0uvrWvKNRwHbl74c/LVkVctxHB/LNub9Cs8KeWR9z0QOm2t6HtPr0i4ob7S79dAIASu/3xrnh8ZeG6L0ea3QdVUbLgX4NKFrjihGpw15Pd8cjy7oLblC5iuNLnjYMXWl/a1qtWOoL0mrd0VV9d75GaVV23niD9olvXXf1jpKSFr16wW+Ft/9zIDwIqebsAAGWQzmj+OWh232HbtFqYEKqwvMtOmzbFFpMGrDcFY9Q/7y8s73LoNi3RZNxihMq7dPVGrOyU+dQ6Lym1bnBZl/sei1j4xPAfb/e5EZtOLV+Zlb3nR0yfVL4yMnl7zouYMbn82wUAKJPBQfrk1vrYd3bzqO0PMDJueKQzm1k5kCtOqAYXPdAevQNmDE9sro8Dtta3GZ6FS3vi/qcLF2hW3gVBei3r7olYsmJkw+DB5U4WPB7xwOIo22Kfdz8S8dBTpdte/3YHzbq/99GIRU+WfrsAAGWyeGVvVuJlIJfJQ3WWLDhifmt20S2MZU+u6o1bFw8at+YJ0hm5qxyWteeiRyWCmiZIr2WplvjAF4AUrF902/Afr7Up4uCdyrf45sTWiIN2KN/28sa3RBy0Y/m3CwAwyou37btlc0xukbZBtZV3mTmhIXbbrGnU9gdKNW7ttUVzTBsn+mJ40oeO3QNys9yaMJ3a5dWklg0u6/KveyOWrx7+4x2yc8S4AZf79vT21UcvlefvEtE0YBXuru6ISzbig4Dn6tCdI1oGnGR29ZRnuwAAZXb5oo7o6F77hrGxvi6rlQ6MbQuf6YkHnh60MOM8xzZj35UPdsTqVMx6jYb6ujjcrHSGKYXm1z1SuEDzM4NKY1FbBOm1qq0zYnnbyM6qHlzW5br7+hYzLZXB27v6noiVhZ8+l6WczLX3rPvfEgCgCrR15bJQYiC1lKE6Z+4+b05LjGtyxQljW0d3xBWLBo1bPiRiBBcdXdWZK5hkQG0RpNeqwbPRl63um5E+XJtOidhjbvnKnczeJGKnLcu/2Ofm0yJ22br82wUAqJBFR+dv0hRzpzaM2v4ApSlZ0NJYFwfP8UEZ1VfXes60xpg/fcDV7DAEaUb68vbCWehmpdcuQXotSqtYP7Gs8LaLb43o7h252eEr2iKuuSfKNiv8mZUR199fuu1taLvLVvXNvAcAqFK3LO6Kp1b1FNz2AouOQlWULLh+UMkCCzNSDdJC2Y+vKBy3jtzWuMXwpKjskoUd6wTpuZStUXME6bUozT7v6Brhsi67F/6caoanxUtLIS0nP3h7qRZ7qsleSukqx/Vtd2M+gAAAqHBpwuo/Hyh8A3n4vNbslAyorvIuu27WHLMmigkY21K8eeEDhX37sLkt0ahrM0LlXdJSeanEC7XHy0gtGlzWZeHjEfcvHv7j7bxVxBbTy1fWZY9tImZMLt/28nabG7Hp1EHbvaX02wUAqLDyLtPG1cfeWwxYZB4Yk/71cGcs7yicGOSKE6px3JrcWh/7zjZuMTz3P90di54pXKBZeZfaJEivNWnW9lPLRzYMfuGgcicPPhlxz6NRtvIq9z8WsfCJ0m1vQ9td8HjEAxvxAQQAwBjxyPKeuOvJwisajxS2wZiXLq69dMG6CzO64ISx7vGVvXHb44NKFynvwkYYfJVDKo/VO2CdCWqDIL3WpBB94IGegvVUnmS4WhojDt2lfLPDxzVHHLRj+baX19oUcfBO5d8uAECFloA4YKvmmNgsboNqO7ZnTWqIXWY1jdr+wEi5cNCio2lG+uQW4xbDc/EDHdEzIE9L/0xhOrVFkF7rZV3SAp1poc7hSqH2+JaRC+afzaE794XaeakO+0W3RcmlED2F+OV6ngAAFeayhR3R2bP2DWNTQ10cOtfChDDW3bekOx5cWliy4AXzHduMfVcs6oiO7rXjVmN9XRy2jVnpDE8q5XLjo4VXOSjvUnsE6bWkvatvodGBLrhpZMud3PhAxJIVG/eYRbe3Z+HP/7o3Yvmg51SO7V53X8TSVaXfLgBAhUiLal3z0KASEMq7QFW4YNCs9EPmtGYXH8NY1taViysXDRq3tvUhESN3lcPKzlzBJAOqn6GxlqQzoV23jrjvsYgV7RE9PRFTJkQcu/fwH/OWhRGd3RG7bB0xaVxf6ZiNebxnc/19EavaI3bZKmLiuIglK0u7vYEfEKTtbrlJ38Kq6cOCcmwXnos1V2m0NtTFkWYPAVBCS1b3zbxa2tYT1z/SGYuWdht7oAp09+SiN5fLgsdbF3fGA093xyFzHNuMfU+39fSPWzc8atxi4zTURbR35yKXy0VXTy52ntUUTaYo15S6XGp9ak8qibK6I2Ly+JF5vNSN0szwFG43lOFVJG0vza5P+19fphpni5+J+M+fRRyzV98HEFAhVu49J3LjmqOzqzduetpJIQClld4wPrmqNy5e0B6Hb9MaU8d5BwnVwLFNtdK3GUmN9RFLVvXGomXd8do9J8SsiQ2jvUuUkRnptaqxYeRC9KSurrzhctre1FEIs+vWhPjWJ6GC1HV0R3dXT3Q3NWeHBgCUUio3m8abNORk3409UBUc21QrfZuRlFVy0YdqliAdhmTNqGvkpYJMuPKeWHDdw/HoK4+Iuplq1QJQevkzoSyUGOV9AUaOY5tqpW8zkvSh2iVIh6HwETaVqK5O1wSgrPJjjrEHqotjm2qlbzOS9KHaJUiHIVnz+bVXTSpKX380uwKAcjGzD6qTY5tqpW8zkvSh2iVIh6Ew7ZdKnZFudgUAZaTWLFQnxzbVSt9mJOlDtUuQDkN9tUwLfNd71aSC5Bedd1IIQLnkp/QZe6C6OLapVvo2I0kfqlmCdBgq14JRabLZFWv/BwClNnDcMfZA9XBsU630bUaSPlS7BOkwFK4Fo4KvU9Q1ASiX/Hhj7IHq4timWunbjCR9qHYJ0mFIpJVUaI30NRVe8lVeAKCU0nhj7IHq49imWunbjCR9qHYJ0mEozEingoN0XROAchm4yLWxB6qHY5tqpW8zkvSh2iVIhyEvNrrmCypF6o9OCgEoI5fIQ3VybFOt9G1Gkj5UuwTpMGRGXirNmhnp1sEFoEwGjjvGHqgejm2qlb7NSNKHapcgHYbCyEslMrsCgDLLxhtXQ0HVcWxTrfRtRpI+VLsE6TAkRl4quEb6mv8BQKnlRxxjD1QXxzbVSt9mJOlDtUuQDkORLfOtRjqVWSM98RkPAOVm7IHq5NimWunbwHAJ0mEoLPNNJc9I1zUBGIVTImMPVA/HNtVK32Yk6UO1S5AOQ2XkpUKD9Po1XwBQamm8MfZA9XFsU630bUaSPlS7BOkwFBYbpRJZOAeAMsuPN8YeqC6ObaqVvs1I0odqlyAdhmRNLWqvmlRit/QZDwBlMnDcMfZA9XBsU630bUaSPlS7BOkwFPkQ3WKjVJI1H+yYXQFAuWTjjauhoOo4tqlW+jYjSR+qXYJ0GBKfYVOJ1iw2qmcCUCZm9kF1cmxTrfRtRpI+VLsE6TAU2eokZqRTYVJ/NLsCgDJSaxaqk2ObaqVvM5L0odolSIehyI+6XjWpJHVpRvra/wFAqfXN6MtfEWXsgWrh2KZa6duMJD2odgnSYahcC0alMbsCgHIbcH28sQeqiGObaqVvM5L0oZolSIehUNqFSlRf1x+i65oAlLGqmNMiqDKObaqVvs1I0odqlyAdhkJpFyqR/ggAAABQUoJ0GCpBOhVZI73vU/GcrglAuWb2rZnVZ1YWVA/HNtVK32Yk6UO1S5AOQ2FGOhUcpCvfD0C5DBx3jD1QPRzbVCt9m5GkD9UuQToMlSCdiuyTuiYA5ZMfb4w9UF0c21QrfZuRpA/VLkE6DEX+xdKLJhWmrv+kUOcEoPTSeJMfd4w9UD0c21QrfZuRpA/VLkE6DEVWVE1BLCpM6pMuUwRglOYWGHugeji2qVb6NiNJH6pdgnQYKteCUWnysytcYgZAmQwcd4w9UD0c21QrfZuRpA/VLkE6DIXFRqlEa/qj2RUAjAZjD1QnxzbVSt8GhkuQDkMlSKcSZ6TnKw/pmgCUqaqYindQfRzbVCt9mxJUV6UGCdJhKMxIpxL1L5yjawJQHi6Rh+rk2KZa6duMJH2odgnSYSjq1nx87SNsKsma/qi0CwDlYtE2qE6ObaqVvs1I0odqlyAdhsrIS6XJZlekWel9XwBQavkxx9gD1cWxTbXStxlJ+lDtEqTDUGQvlq4Fo9KYkQ5AeZnZB9XJsU210rcZSfpQ7RKkw1ApqkalUSMdgDJTaxaqk2ObaqVvM5L0odolSIehyJb5ViOdCq2R7qQQgFG4SM/YA9XDsU210rcZSfpQ7RKkw1D4CJtKnZHuMkUAymjguGPsgerh2KZa6duMJH2odgnSYaiMvFQasysAKLP8eGPsgeri2KZa6duMJH2odgnSYSgUoqaCZ6Rn/xzlXQGgdhh7oDo5tqlW+jawsQTpMNzyLlAp1iyck0ql53RNAMogjTmWjoHq49imWunbjCR9qHYJ0mEozEingpegzy+gAwCl5hJ5qE6ObaqVvs1I0odqlyAdhjMb3ceP1e2nF0V88hcRe24Tce7H1/39nLcX/jyuOWL2JhEnHBDxjhdGjGuJ0VpstFzOvaMtPnXB8uzfP3rVtNh7i+aC3+dyuTjmR0/F4yt749C5zfH/Xjotu32PMx+Pk3YfF/95+OQy7i0A1bBoW37s+eVJ02OXWU3Dfpy2rlz8+PpVse+WzbHfloXjF9S6sXxsQ6UuNvrrW1bH/1y8Inad1Ri/OGmTdX6f3iMNNKGpLnac2Rhv3mdCPH+bMr+35DmRCNUuQToMlVm/1e8PV0dsNSPipgURix6PmDtr3fscunPEKw/q+/eqjoh/3RvxlT9E3PlQxHdOKe/+9s+uKN/0ivxmWhoizru7PfaZXXiCd90jnVmI3tyQdq+ub9/6d7fwZwDGnr4xp+97uV7T184m3LhtdvTk4jvXrop310Xsv5WAAqrl2IZK69t5f727PbaY3BC3Pd4dDy3ria2nrhvFHbR1c7xkp3GRy0U8tqInfnPL6nj/n5bGt142LQ6eY6yqNF6rapcgHYZCaZfq9+CTEdffH/G990Z87KcRf7gm4t9PWPd+8zaLeOXz1v78xiMiurojzrshoqM7orWpqmek5x06tyXOv689Pnb45GgccKVGOlncedPGWNrW27eLg/7OEQQwto3mzL6N3ebAxeaMR1A9xzZUYt9+eFl33PRYV3z9xVPjv/65PP56V3uccuDEde43Z2pjvGTHcf0/v3Db1jjhZ0/FL25cFYcI0iuO16raJUiHoRKkV7cUnE+ZEHHUHhGX7xvx+2siTn3Z+u87uB/MnNp3W1NDeftIfnZF9u8ybXLN9+N2HBcX3t8RVz/YGYeuueywqycX59/bHu86YGJ24tdfv33g7jqEAMa00ag127+ZDWwzjT/fvXZlXLqgIx5c2hM9vRE7bdoY7ztoYv/M80eWdWelx5JvX7Mq+0pOOWBCvPegSeV5IlDBKvHYTu58oiu+ccWKuPGxrujNRey+WVO8/3kTY4/N15Zn+sPtq+MT5y+Pn544PS64tz3+dFdbtHdFHDSnOU4/ckpMH19fludDZRqtGulpgtHklro4bF5LvPChlvjL3W3xnoPWDdIH79f8TRpj2ri6eHh5j/dOFUib1C5BOgyFGenV7/dXRxy3d0RLU8TLDuyrl37zgog95xXer7Mr4pmVff9evaa0yzlX9P1NU5lfWvMz0ss5vWLNdmZPbog9Nm+K8+5pi+fP6wspLl/UESs7c3HcDq3xi5tWrd23wl0GYAzLv7av7zW+pBsdsO3BVnXl4re3tWXjz6t2bcx+/t1tq+Odv38mfnXyJrHTpk0xfUJ9fOoFk7NZgUdt2xJHbdua/e32MxqNTVChx/Z9T3XFG89+OiY218Vb950QjfURZ9/SFm855+n4yaunx+75MH3N355x8fKY3FKfzfp9dHlP/OyG1fE/DcvjKy+eWqYnRCUalb4dkQXnL9yuNZob6+LFO4yLX9/SFrc93hW7bbbuFcwD92tFR28sb8/FVlPqjU8VSJPULkE6DPXVMpWvsNhodbp5YcR9j0V87vV9bXzg9hFbTO8L1/eeX3jfsy7r+xro2L0jvvKW8vePNdsr51YHXhZ//I6t8bUrVkZHdy5aG+viz3e2Z4u3zZrYsM798/92BAGMbaNRHuXZtjmlpS7Of9vMaG5Y+9tX7zouXvyTp+KXN62O/z56Skxoqo9jtm/NgvQUnr90p7WX0QOVeWyfeeXK6O7Nxc9P3CS2WlNb+mU7jYvjfvJUfOWyFfHTE/sWb8z/7dTW+vj+K6b11zBONad/fuPqWNnRG5NazEqvVaPRt29/vCseeLon/vPw1myb+8xuis0m1sdf7mrLrqoYqLMnl5XFzNdI/8aVK6InF3H0dn1/S2XRJrVLkA5DYkZ6Vfv9VREzp0QcsvPadn7p/hG/vSriM6+NaKgvDM3felTfv9s61tRV/3vEe78b8f33lb20S7kvlhh4aeRxO4yLz1+yIruU/pC5zXHJgvb4+BGTCz5PGPzZgs+iAMa2gbP66kdh7FnfNusb6iIfS/TmctlMvlxE7DqrKSsLkf+b/Pf8/Aigco/tnt5cXLmoM46a3xpzpq2NL2ZNasgmc5x9a1us7uyNiS1rZ+2etPv4aBjwQPvObo6f3LA6Fq/oiSmtgvRaNRp9+893tcWM8fXZQqLZNuvq4kU7tMaf7myPjx42qaCfpiuq0ldeU33E2/edEG/dd7yxqgKJhGqXIB2GdT2YV82qkwqppvroB+8U8VBf7dTMPttGfOdvEZffEXH4bmtvTzPVD9t17c/H7hMxfVLE6WdFnH9zxDF7lbm0S9//yvXZeN+2+ra2yfiGOGjrlvjznakOZS6bOXHsduOicK/W7tfgnwEYe9a+xudf6cuzzWcbR35/++r44fWrYsHT3dHVt951ZsspDQP+Jv845dt3GCsq7dh+pq032rpzsc20xnV+N396U/Tm2mLxit7YrmXtMb7FpIHHe/SH5+nDNcd87Sp3304fAqX66Ads1RyPLFs7IKW6/j+6fnVc/WBXHDJ37SKiR85vidfvOSFb7+PWx7viO9eujPbuXDTU+fCnEnktqV2CdBgqdSmqUwrKH18a8Yer+74G++2VEUcMCNJjPf3g0J37vl99V8Sx5QzS13wrY9/sjyLWfK70kp1a4xP/WBZPre6N589tiSnjBpzwDf7syWdRAGNefswZjQUJN7TNc+9YHR/9+7Ks9vnb95uQfdCbZvGlBUgfWrp2sbb+PzUeQcUf28UWiBz8u/yv00Wkgxe6z2/IMV+7yt23r3moM55c1Rt/ubs9+xosLYZ76DZrg/TNJjXEwWuC9cPnt8a0cfVZGbIDtm6OY7ZThqzSeC2pXYJ0GIr8qOvaqurzuysjZk6O+MKb1v3dn6+L+Ov1ER1dEeMGLGY0uB/09q5dfLR+dEq7lLuYZX67qXbfp85fFjc91hXfOH5q0cVF8xd2AFAdl8iXfUHCDWzz7/e2x1ZTGuJbJ6ytjZx888oV/X+X1K/5rNd4BJV/bG8yvj7GNdbFgme61/ndA093Z6fcm09uKDgPHvw4BbWxHfM1q9x9+493tWX999NHTl7nd/+4tz3Ov6+9b42ppgFXZAzYr5P3HB8/vmFVfP3ylXFMqpOu81YUzVG7BOkwFPkTNK+a1aWtsy8sP+GAvq/BNp8W8burIv5+Y8TLD1xz43rOwP5xU9/3XeeUuUZ6wbdybrL/kJjYXB//ddSUeHh5Txw5v3BBnPXl+44ggLFt8DhQCdtsGDD25v9102OdceOjXbFFCtrW3Da+se9fKzp613mcdNsTK3tj04n1FiWkJlXasd1YX5etwXPBfe3xyLLu2HJKX4Tx1Kqe+NOdbbHP7OaYvOZYfS77nr/9iZU9saIjF1tPbYimAQsUU73K2bdTucsUlr9o+9ZsPanBNpvYEH++qz3+eX97vHjHtb8fuF9N9XXxtn0nxKcvWB4X3tcRL9yutcR7zVB41ahdgnQYEouNVqW/3RCxsj3iRXuvv2332y5ixuSIc66MeMVBfbfd/1jE2Vf0/Xt1Z8R190X86tKIebMiTjqk/IuNlrm0y8Az0fxTfeVu4zd890GX1zqEAMa2gYtcl3vW6jm3tcWlCzvW+XW6/D3NSn/Puc/E4fNa4+Fl3fHLm1bHtjMaY3Vnrn8/xzXXxXabNGaX2m8zvTGrn7z9jMbYYWZT/OO+9viP85bFF140JV6164bHNahWlXhsf+DgSXHFos54zVlL4nV7TYiUe//q5tXR2ZPLFmwcWLol/319pV0GPqcvX7Yifnd7W1zyzpn94TzVrZx9+8IH2mNVZy6O2jbNJF/393vNborp4+vjj3e2xfE79QXp69uvNA5944qV8b1rV8bR2wvSK4n3s7XLiAFDUfbrHCmLFJC3NkUcsfv627ahIeLoPfuC82dW9t128W19X9nv6yNmTY14wxERH391xMQy17Bbs8+jOSP9ud5/Qz8DMLaM5qzVX9y0er2/v/xdm2aB+Vk3r45LFyzLwvKvvnhqnHd3e1z9UGfBfp5xzJQ4/cLl8bmLlkdnT8T7nzcxdpzZNCrPCypJJR7bKUz89cmbxJcuXRHfuXplpGKKe27elB3fe23R/Kz7vr7bHeu1p5xt/sc72qKlMeLQuS0bvILqiHkt2f2Wtq1diHTwfcc11cUb9hof37hyZVzzYEccuPXamuqMLq8btasul8vlRnsnYExY/EzED86PeMuREZtNG+29gbUWPxO5H/8zcvomAGWSyiKk2d6v3XN8bDqxYbR3Bxghjm2qlb7NSPen9MH9a/ecELP0p5piRjoMRdmvc4TKnJEOQG0zmxOqk2ObaqVvM5L0odplBR0AAAAAAChCkA4AAAAAAEUo7QJDpbQLlaauru/SMn0TgHKOPWsWYa8z9kD1cGxTrfRtRpI+VLPMSAcAAAAAgCIE6QAAAAAAUIQgHQAAAAAAihCkAwAAAABAEYJ0AAAAAAAoQpAOAAAAAABFCNIBAAAAAKAIQToAAAAAABRRl8vlcsXuAKzR1R2xZEXEJpMimhpHe29gLX0TgDLr6snF0229MX1cfTQ11I327gAjxLFNtdK3GUn6U+0SpAMAAAAAQBFKuwAAAAAAQBGCdAAAAAAAKEKQDgAAAAAARQjSAQAAAACgCEE6AAAAAAAUIUgHAAAAAIAiBOkAAAAAAFCEIB0AAAAAAIoQpAMAAAAAQBGCdAAAAAAAKEKQDgAAAAAARQjSAQAAAACgCEE6AAAAAAAUIUgHAAAAAIAiBOkAAAAAAFCEIB0AAAAAAIoQpAMAAAAAQBGCdAAAAAAAKEKQDgAAAAAARQjSAQAAAACgCEE6AAAAAAAUIUgHAAAAAIAiBOkAAAAAAFCEIB0AAAAAAIoQpAMAAAAAQBGCdAAAAAAAKEKQDgAAAAAARQjSAQAAAACgCEE6AAAAAAAUIUgHAAAAAIAiBOkAAAAAAFCEIB0AAAAAAIoQpAMAAAAAQBGCdAAAAAAAKEKQDgAAAAAARQjSAQAAAACgCEE6AAAAAAAUIUgHAAAAAIAiBOkAAAAAAFCEIB0AAAAAAIoQpAMAAAAAQBGCdAAAAAAAKEKQDgAAAAAARQjSAQAAAACgCEE6AAAAAAAUIUgHAAAAAIAiBOkAAAAAAFCEIB0AAAAAAIoQpAMAAAAAQBGCdAAAAAAAKEKQDgAAAAAARQjSAQAAAACgCEE6AAAAAAAUIUgHAAAAAIAiBOkAAAAAAFCEIB0AAAAAAIoQpAMAAAAAQBGCdAAAAAAAKEKQDgAAAAAARQjSAQAAAACgCEE6AAAAAAAUIUgHAAAAAIAiBOkAAAAAAFCEIB0AAAAAAIoQpAMAAAAAQBGCdAAAAAAAKEKQDgAAAAAARQjSAQAAAACgCEE6AAAAAAAUIUgHAAAAAIAiBOkAAAAAAFCEIB0AAAAAAIoQpAMAAAAAQBGCdAAAAAAAKEKQDgAAAAAARQjSAQAAAACgCEE6AAAAAAAUIUgHAAAAAIAiBOkAAAAAAFCEIB0AAAAAAIoQpAMAAAAAQBGCdAAAAAAAKEKQDgAAAAAARQjSAQAAAACgCEE6AAAAAAAUIUgHAAAAAIAiBOkAAAAAAFCEIB0AAAAAAIoQpAMAAAAAQBGCdAAAAAAAKEKQDgAAAAAARQjSAQAAAACgCEE6AAAAAAAUIUgHAAAAAIAiBOkAAAAAAFCEIB0AAAAAAIoQpAMAAAAAQBGCdAAAAAAAKEKQDgAAAAAARQjSAQAAAACgCEE6AAAAAAAUIUgHAAAAAIAiBOkAAAAAAFCEIB0AAAAAAIoQpAMAAAAAQBGCdAAAAAAAKEKQDgAAAAAARQjSAQAAAACgCEE6AAAAAAAUIUgHAAAAAIAiBOkAAAAAAFCEIB0AAAAAAIoQpAMAAAAAQBGCdAAAAAAAKEKQDgAAAAAARQjSAQAAAACgCEE6AAAAAAAUIUgHAAAAAIAiBOkAAAAAAFCEIB0AAAAAAIoQpAMAAAAAQBGCdAAAAAAAKEKQDgAAAAAARQjSAQAAAACgCEE6AAAAAAAUIUgHAAAAAIAiBOkAAAAAAFCEIB0AAAAAAIoQpAMAAAAAQBGCdAAAAAAAKEKQDgAAAAAARQjSAQAAAACgCEE6AAAAAADEhv1/C+qhu7IqVkQAAAAASUVORK5CYII=" }, "metadata": {}, "output_type": "display_data" } ], - "execution_count": 15 + "execution_count": 17 }, { "metadata": {}, @@ -1124,8 +1330,8 @@ { "metadata": { "ExecuteTime": { - "end_time": "2025-01-10T14:03:05.650284Z", - "start_time": "2025-01-10T14:03:05.111943Z" + "end_time": "2025-01-14T16:33:30.739716Z", + "start_time": "2025-01-14T16:33:30.267097Z" } }, "cell_type": "code", @@ -1135,15 +1341,15 @@ { "data": { "text/plain": [ - "
" + "
" ], - "image/png": "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" + "image/png": "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" }, "metadata": {}, "output_type": "display_data" } ], - "execution_count": 16 + "execution_count": 18 }, { "metadata": {}, @@ -1161,8 +1367,8 @@ { "metadata": { "ExecuteTime": { - "end_time": "2025-01-10T14:03:19.687073Z", - "start_time": "2025-01-10T14:03:19.327787Z" + "end_time": "2025-01-14T16:33:31.057793Z", + "start_time": "2025-01-14T16:33:30.740716Z" } }, "cell_type": "code", @@ -1176,21 +1382,21 @@ { "data": { "text/plain": [ - "
" + "
" ], - "image/png": "iVBORw0KGgoAAAANSUhEUgAABiIAAAFqCAYAAACXjkI0AAAAOXRFWHRTb2Z0d2FyZQBNYXRwbG90bGliIHZlcnNpb24zLjkuMiwgaHR0cHM6Ly9tYXRwbG90bGliLm9yZy8hTgPZAAAACXBIWXMAAA9hAAAPYQGoP6dpAABaAUlEQVR4nO3dd5xU9b3/8feZ2V7YwtKWskvvHakiqCg2BEs0pliTm3jNvXpTrze/JCa5icb0xETjtRujEY0ggg0EFOkIKlV6XUDYZftsm/P742w7W2B32TNn5uzr6WNd5jtnZj5zvt/vnpnzOd/v1zBN0xQAAAAAAAAAAIADfG4HAAAAAAAAAAAAvItEBAAAAAAAAAAAcAyJCAAAAAAAAAAA4BgSEQAAAAAAAAAAwDEkIgAAAAAAAAAAgGNIRAAAAAAAAAAAAMeQiAAAAAAAAAAAAI4hEQEAAAAAAAAAABxDIgIAAAAAAAAAADiGRAQAAAAAAAAAAHAMiQgAAAAAAAAAAOAYEhEAAAAAAAAAAMAxJCIAAAAAAAAAAIBjSEQAAAAAAAAAAADHkIgAAADSGxukwlK3owAAAAAAAB5EIgIAgI7uWK50x5+k6x8kGQEAAAAAANodiQgAQHh7YYXU6YvSB9vOve2Ib0lX/bTtr/XL+dZrHTzZ9ueIRJnp0rP3Spv3STc8JBUFHH/JNYfKlPWrY5r/aYnjrwUAAAAAANwV5XYAAADAYb+c37LtxvWX1u6SbnhQWvBDKT7G2bgAAAAAAECHQCICAOAdm34vGYbbUYSfh15t3fYb9kif50t9ujgTDwAAAAAA6FBIRAAAvCM22u0IWqewVEqOd/51Cl469zYlZdLND0sf7pSe+g+SEM2oCpoqrzIVH83slgAAAAAAtBSJCABAZAia0p8WSU+8ay2u3DtD+u510pdn1G0z4lvWCfQlP7E/9ol3pL++KR36XOqVId19hZQUJ939mLT4R9L04fbtyyuln74ovfiBdKpAGpQp/eQWafbYxnG9ulr629vS1oNSVVAa1lu6d440b7J9u05flL50kfTF6dIvX5E+PSCN7dc4Vjc0TEI0jD1UYZQH9ec1RXpjZ6mOF1YpJc6n6dmx+s70ZPVKsT6ylFWaGvnHHF0zJF6/uzqt9rH3v3VG//i4RHeMT9QDs1Jqy+9ZmKsV+8r08b3dFeWzRssUlAX1lzVFenNXqXIKq5QU49OF2bH63kXJ6pNa99Fo/qcl+u6SM3rh5s7adLRcr2wt0bGCKj10Raq+MDIhRHsFAAAAAIDIRyICABAZfvqSFCiX7pwlxURJTy6V7n5U6t9dmjy4+cf9fqH0kxelMX2lB26xTrr/6Q0po1Pzj/nGX6Vov/Qf10gVlVYS40u/kT76vZTVtW67n/1T+s1r0qzR0g9vknyG9MYG6dY/SL+5Q/q32fbn3bxPen29dNslVlIiXJRVWAtUu5iEqKgy9dWXc7XxaLmuGhynr1+QpAN5lfr75mJ9cKBMi27toh6d/IqNMjS+Z4zWHCy3Pf7Dg2XyGdLqg2W1ZaZpau2hcl3QK8aWhLj++VM6Vlilm0YmaFBGlE4WBfX85mLNfa5Mi27LqE161PjF8nxVVEm3jE5QUoxP/dL5+AQAAAAAQGvwTRoAEBnKK6UVv7SSEJJ1wnzUf1qjEZpLROQWSQ++Ig3vI73zUymuevHl2y6Rxv9X86/VOVl6+ft1601MHy5d/EPp6WVWMkOStuy3khDfmWuNlqhx95XSLb+xEie3XGSfemnHEWnhD6WLR7ZtHzglLUla9nPJ5950Q698WqKNR8v1jYmJ+p+L60Y0TMuO1Z2v5OpX7xfoD9dYIyCm9onV6oOF2p9bqb7pUTpaUKmDZ6p03fB4vbatVJ8XV6lLol+7TlXqVElQU7Nia5/vdx8U6lB+pRZ8tYuGda2byuvGkfGa/dTn+v2qQv223kgLSQpUmlpyexemYwIAAAAAoI34Rg0AiAxfu6wuCSFJmenSgB7SvuPNP2b5J1KgQrprVl0SQpK6pUpfuLD5x919pX3R6/H9ramc9ubUlb28ytrmSzOk0wX2n6vGW+s/rP/M/rwjs1xNQpRWBLV4Z2nTd7qYhJCkt3YH5DOke6Yk28ov7R+nYV2j9O7ugIKmKUm1iYXVh6zRD6sPlstvSP81LVmG6kZFrKn+XbO9aZpasL1Uk3rFqnuST7klVbU/CdGGxmbG6P0DZWroK2MSSUIAAAAAAHAeGBEBAIgM2V0bl6UnSYdPNf+Yg59bvwdmNr5vYI/Wv1ZuUd3tXUcl05TGf7v55zmZb7894Cyv6bBAhak7X83VukPlGtIlWv07h9dHgMNnqtQtyaeUuMYn/AdlRGv7yUrllgSVkejX6B7RSooxtPpgmb48JlGrD5ZpVPdoZaVFaUiXKK0+WK65wxK0+lC5UuMMDe9qvdfTJUHllQb1/oEyjf3ziSbj8BmNy/oyFRMAAAAAAOeFb9YAgMjgb+aK9Oqr5EP+WqZpjYh49b+b335oL/vt+Jimt3OYlYQ4rXWHyvXbq1PDLgnRWlE+Qxf0itGaQ+UyTVOrD5bphhHW4tFTs2L1TvXoibWHyjQ1K1ZG9eiWmtq7MCtGd09ObubZG4uPbiI7AQAAAAAAWiyyz0QAAHA2fbpYv3cfk2aMsN+3O6fx9q3Rv4e09GOpd4Y0uOf5PZeDyqtM3fXqaX14sFwju0frQF6lfr+q4JyPu2dKsmL8oTsB3yfVr5X7K5UfCDYaFbH7dKWSYwylJ9SVT82K1fJ9ZVqyK6DjRXXrQEzLitWTG4v15q6ACspM2/oQnRN86hRrqKjc1IXZsQIAAAAAAKFBIgIA4F0Xj5Rio6Unl0pfmVm3TsSJM9L8Vef33F+cLv3tLWtR6uf/q/GoiJNnpK6p5/ca7aAgENSGI+WSpE+PV+jT4xUtety/TUwKaSJi9sA4Ld9XpkfXFum/Z3aqLV++N6BtJyp03fB4+eqt21GTYPjdqkLF+qUJvay6ndg7Rn5D+v2qQmu7PnWjUHyGoXnD4/XcRyVavLNUVw+pt5B4tVPFVcpI9DvyHgEAAAAA6KhIRAAAvKtzsvTfN1jJgst/It10oVRaLj2zzBrRsHmffVHq1hjfX7r/RunBV6RpP5DmTZZ6pEnH86Qt+6V3NkunX2jf99MGGYl+PXpduu5+LVfDukbruZs7q1Ns+C28fOPIBL2ytVSPrivSkfxKTewdqwN5lfr75mJ1SfTp+xd1sm0/vGuUUuMM7Tldqcl9YhQXZdVjcqxPo7pHa3NOhbom+TQwI9r2uO9d1Ekbj5TrnoV5enNXqcZmxijab+hoQZWW7w1oZPdo/fbqtJC9bwAAAAAAOgISEQAAb/vOPCk5Xnr0TemBF6VeGdJ/XmMtGLB5X90oiba4/0ZpbD/psbekvy6RSsqkLp2kob2lh29vpzdw/i7tH1ebjLj1n6fDMhkR7Tf0/E3p+vOaIi3aUaq3PguoU5xPVw2J13enJyuzk32UgmEYmtwnVm99FtDUPvZplqZmx2pzToWm9Gk8/VKnWJ/+9ZUMPb6+WIt3luqdPQFF+Qx1T/Lrgl4x+uLoBEffJwAAAAAAHZFhmk6s8gkAQJj77tPS429Lux+TuqW6HU1ILNsb0G/eL9BzN3VWF6YfAgAAAAAAIUIiAgDgbYHyxqMejudJE75tjY5Y+2t34nJJ0DRtay0AAAAAAAA4jamZAADe9sF26UcvSHMmSj3TpUOfS8+8JxUFpAducTu6kCMJAQAAAAAAQo1EBADA2/p1l/p2k559T8otlOKirXUdvj1Punik29EBAAAAAAB4HlMzAQAAAAAAAAAAx/jcDgAAAAAAAAAAAHgXiQgAAAAAAAAAAOAYEhEAAAAAAAAAAMAxJCIAAAAAAAAAAIBjSEQAAAAAAAAAAADHkIgAAAAAAAAAAACOIREBAAAAAAAAAAAcQyICAAAAAAAAAAA4hkQEAAAecPToUT355JP6xS9+oQceeEDHjx+XJC1evFjPPfdcq59vz549+uUvf6ni4uL2DhUAALSzFStW6IEHHlBJSYnbobimZh8AAIDwRCICAIAIV1VVpfnz56u0tFRXXHGFrr/+eqWkpCgvL08fffSRpk+f3urnHDBggNLT07Vq1SoHIgYAAAAAAB0JiQgAACJcXl6ezpw5o6lTp2r8+PEaNWqU4uPjtW7dOqWmpqpv375tet7x48dr48aNKisra+eIAQAAAABAR0IiAgCACFczfVJcXFxtWVVVlT755BMNHz68zc87bNgwVVVVafv27ecdIwAAAAAA6Lii3A4AAAC03YIFC7RlyxZJ0ssvvyxJys7O1owZM1RSUqJ+/frZtn/ttde0bds2feMb31CXLl1qy59//nkdPXpU99xzj5KTkyVJiYmJ6tatm3bu3KmxY8eG5g0BAIA2Kykp0eLFi7Vnzx75fD6NGjVKl112maKi6r76b968WZ988olOnjypQCCg9PR0TZw4URdccIHtuY4dO6Zly5YpJydH5eXlSkpKUt++fTV37tzabUzT1Lp167Rp0ybl5eUpNjZWQ4YM0axZsxQfH99snKtXr9Y777yj++67T6mpqbb7li5dqjVr1ui73/2u4uPjdfDgQa1bt05Hjx5VUVGREhMTNWzYMF166aWKjo5u9jXOnDmjP/zhD5o3b57GjBlju++BBx7QzJkzNXPmzNqygoICLV++XJ999lntfpk6dSqfgQAAaCckIgAAiGDjx49XcnKyPvjgA02aNEk9e/ZUYmKiDh8+LMMw1KNHD9v2V155pfbv368FCxborrvuks/n08aNG7V3715df/31tUmIGj169NDOnTtD+ZYAAEAbzZ8/X6mpqbr00kt15MgRrVu3ToFAQNddd13tNhs3blSXLl00ePBg+Xw+7dq1S4sXL5Zpmpo4caIka7Tl888/r4SEBF144YWKi4vTmTNntGPHDtvrLVq0SFu2bNHYsWM1adIknTlzRuvXr9fx48d15513yu/3Nxnn8OHD9e6772rbtm2aNm2a7b5t27apf//+tYmM7du3q6KiQhMmTFBCQoKOHj2q9evXq6CgQDfddFO77LeioiI98cQTMgxDEydOVGJionbv3q2FCxeqrKxMkydPbpfXAQCgIyMRAQBABOvdu7eqqqr0wQcfKCsrS8OGDZMkffzxx4qPj1dsbKxt+7i4OM2dO1fPP/+8Vq1apZEjR+qdd97RkCFDNGrUqEbPn5aWppKSEhUXFysxMTEk7wkAALRNamqqbrnlFknSxIkTFRsbqw0bNmjq1Knq1q2bJOn222+3jSSYOHGi/v73v2vNmjW1iYjDhw+rtLRUX/3qV5WZmVm77SWXXFL770OHDumjjz7SDTfcoJEjR9aWZ2dn6+9//7u2b99uK68vJSVFvXr1apSIOHr0qPLy8mwjFWbNmmWLd/z48UpPT9eyZcuUn5+vlJSUtuwqm/fee0+maeqb3/ymEhISJEkTJkzQK6+8ohUrVmj8+PFnHX0BAADOjTUiAADwoNLSUtuaEfX1799fEyZM0MqVK/XPf/5TUVFRmjNnTpPb1lyNWFJS4lisAACgfdQkEmpMmjRJkrR79+7asvon1AOBgEpKSpSVlaW8vDwFAgFJdetOffbZZ6qqqmrytbZt26a4uDj169dPJSUltT+ZmZmKiYnR/v37zxrr8OHDdezYMeXm5tqeMyoqSkOGDGky3vLycpWUlKh3794yTVM5OTlnfY2WME1T27dv16BBgyTJ9l4GDBigQCDQLq8DAEBHx4gIAAA6oMsvv1w7d+7U8ePHdcMNNzQ72sE0zRBHBgAA2io9Pd12Oy0tTYZh6MyZM7Vlhw4d0ooVK3T48GFVVFTYti8rK1NcXFztKMsVK1ZozZo1ys7O1pAhQzRy5Mja9SZyc3MVCAT061//uslYiouLzxrr8OHD9fbbb2vbtm2aPn26TNPUtm3bNGDAANuIzvz8fC1fvly7du1SaWlpo3jPV0lJiQKBgDZt2qRNmza16b0AAIBzIxEBAIAHxcfHN/qyXl9OTk7tl+qTJ082u13NlZE10xQAAIDIYRiG7XZubq6ee+45ZWRkaPbs2UpJSZHf79fu3bu1Zs2a2gsQDMPQTTfdpCNHjmjXrl3au3evFi5cqDVr1uhrX/uaYmJiZJqmEhMTdcMNNzT52uf67JCcnKysrKzaRMSRI0eUn5+vyy67rHabYDCo5557TqWlpZo2bZoyMjIUExOjgoICLViwoE0XTASDQdvtmucYNWpUo0Wta9RMawUAANqORAQAAB6UkZGhTz/9VIFAoNEUTeXl5Vq4cKG6dOmi3r1768MPP9SQIUPUs2fPRs+Tl5enhIQE1ocAACAC5ObmKi0tzXbbNE2lpqZKsqZaqqys1C233GJbW6G5aZR69eqlXr166dJLL9Wnn36qV199VVu3btW4ceOUlpamffv2qXfv3m1eP2H48OFavHixTp06pW3btik6Orp2iiTJulji9OnTuu666zR69Oja8r17957zuWuml6y5qKJGfn6+7XZCQoJiY2Nlmqb69evXpvcBAADOjTUiAADwoLPNnbx06VLl5+fruuuu0+zZs5WamqoFCxaosrKy0bY5OTnq3bt3KEIGAADnaf369bbb69atkyQNGDBAUt0IifojCQKBgLZs2WJ7XGlpaaPRBt27d5ek2s8Lw4cPVzAY1Pvvv98ojmAw2CgB0JRhw4bJ5/Np69at2rZtmwYNGqSYmJja+5uK1zTN2vd1NrGxsUpISNDBgwdt5Rs2bLDd9vl8Gjp0qLZv397kKFGmZQIAoH0wIgIAAA/q06ePEhIStG/fPvXt27e2fP/+/dqwYYNmzJihHj16SJLmzp2rZ555RsuXL7dNh1BcXKwTJ07oggsuCHn8AACg9c6cOaMXX3xRAwYM0OHDh/XJJ59o5MiRtUmE/v37y+/368UXX9T48eNVXl6ujz76SImJiSosLKx9no8//lgbNmzQkCFDlJ6errKyMn300UeKjY3VwIEDJUnZ2dmaMGGCPvjgAx0/flz9+/eXz+dTbm6utm3bpiuvvFLDhg07a7yJiYnKzs7WmjVrVFZWphEjRtjuz8jIUHp6ut555x0VFhYqNjZW27dvb1GSQ5LGjRunVatW6fXXX1dmZqYOHjyo06dPN9pu1qxZOnDggP7v//5P48ePV5cuXVRaWqqcnBzt27dPP/jBD1r0egAAoHmMiAAAwIP8fr9Gjhypbdu21ZaVlZVp4cKF6t69uy666KLa8qysLE2ePFmrV6/WkSNHast37Nghv9+v4cOHhzR2AADQNjfeeKP8fr+WLl2q3bt3a+LEiZo7d27t/RkZGbrpppskSe+88442btyo8ePHa9KkSbbnycrKUmZmprZu3ao333xTH374odLT03XbbbfZpn665pprNGfOHBUXF2vZsmVatmyZ9u/fr1GjRrV4ROWIESNUVlZmS3LU8Pv9uuWWW9S9e3d98MEHWrFihTp37qzrrruuRc89Y8YMjRs3Ttu3b9e7776rYDCoL3/5y422S0pK0te//nWNHTtWO3bs0JIlS7R27VqVlpZq1qxZLXotAABwdobZltWdAABA2MvLy9MjjzyiL3/5y22a8/ixxx5Tdna2rrjiCgeiAwAAAAAAHQUjIgAA8Ki0tDSNHTtWq1atavVj9+zZo9zcXE2fPt2ByAAAAAAAQEfCiAgAAAAAAAAAAOAYRkQAAAAAAAAAAADHkIgAAAAAAAAAAACOIREBAAAAAAAAAAAcQyICAAAAAAAAAAA4hkQEAAAAAAAAAABwDIkIAAAAAAAAAADgGBIRAAAAAAAAAADAMSQiAAAAAAAAAACAY0hEAAAAAAAAAAAAx5CIAAAAAAAAAAAAjiERAQAAAAAAAAAAHEMiAgCAMDZ//nx985vf1IQJExQbGyvDMGp/2qK8vFyPPvqoLrnkEnXt2lXR0dGKi4tTVlaWrr/+ei1atOicz3Hw4EF16tTJFsszzzzTpngAAAAAAID3GaZpmm4HAQAAmjZmzBh9/PHHTd7X2kN4ZWWlLrvsMq1YseKs2/3P//yPfvGLXzT7mpdddpmWLVtmK3/66ad1++23tyoeAAAAAADQMTAiAgCAMGYYhvr376+bb75ZM2bMOK/neu2112xJiHHjxulnP/uZvv3tbyslJaW2/OGHH1Z+fn6Tz/HYY481SkIAAAAAAACcTZTbAQAAgOatXr1a8fHxkqQHHnhAK1eubPNz7d2713b7nXfeUefOnSVJmZmZ+u53vyvJGjlx5swZW3JCkg4cOKDvf//7kqR58+ZpwYIFbY4FAAAAAAB0HIyIAAAgjNUkIdrDsGHDbLdffvlllZaWKicnR0uXLq0tHzp0qPr06WPb1jRN3XnnnSoqKtKgQYP0y1/+st3iAgAAAAAA3kYiAgCADmLOnDmaN29e7e1///d/V0JCgjIzM/XWW29Jki655BK98cYbjRbD/utf/6rly5fL5/PpmWeeadcECQAAAAAA8DYSEQAAdBCGYehf//qXfvSjHzVKNEhSVlaWvvKVr6hfv3628n379ukHP/iBJOk73/mOpkyZEpJ4AQAAAACAN7BGBAAAHURFRYVuvfVWvfTSS5KsqZpuvPFG5ebm6qmnntLBgwd15513avPmzfrTn/4kqW5KpuLiYg0dOlQ///nP3XwLAAAAAAAgApGIAACgg/jb3/5Wm4RITU3V6tWraxekvuCCC3TbbbdJkh555BF961vf0qBBg/TSSy9p5cqV8vv9evbZZxUbG+ta/AAAAAAAIDIxNRMAAB3EsmXLav89aNCg2iSEJE2YMKH236Zp6pNPPpEknThxQpJUVVWliRMnyjAMGYahvn372p77jjvukGEYeuaZZxx8BwAAAAAAIBKRiAAAwEOeeeaZ2mRBw3Ugqqqqav/92WefKT8/v/b2xo0bbduyGDUAAAAAAGgvTM0EAEAYe/TRR7V3715J0urVq233ffe736399913363+/fuf9blmzpypRYsWSZLOnDmjqVOn6sYbb1ReXp6eeuqp2u0SExM1bdo0SdLAgQN1ww03NHqukpISvfnmm7W3J0yYoKysLGVnZ7fuDQIAAAAAAM8zTNM03Q4CAAA0bebMmVq5cuU5t1u+fLlmzpypZ555RnfccUdtef3DfGlpqS655BKtXbu22efx+Xx64oknbM/RlAMHDtimZ3r66ad1++23nzNOAADQsRUXF8s0TRmGocTERLfDAQAAIcLUTAAAdBDx8fFauXKlHnnkEc2cOVMZGRmKiopSXFyc+vXrp69+9atau3btOZMQAAAAbWWaZu0PAADoOBgRAQAAAAAAQqKoqKh2RERSUpLb4QAAgBBhRAQAAAAAAAAAAHAMiQgAAAAAAAAAAOAYEhEAAAAAAAAAAMAxJCIAAAAAAAAAAIBjSEQAAAAAAAAAAADHkIgAAAAAAAAAAACOIREBAAAAAAAAAAAcQyICAAAAAAAAAAA4hkQEAAAAAAAAAABwTJTbAQAAgNAqKSmRaZoyDEMJCQluhwMAAAAAADyORAQAAB1MMBisTUQAAAAAAAA4jamZAAAAAAAAAACAY0hEwPPef/99zZkzR5mZmTIMQwsWLDjr9rfffrsMw2j0M3z48NAEDHhIa/ufJL3wwgsaPXq0EhIS1KNHD9155506ffq088ECHtKWvveXv/xFQ4cOVXx8vAYPHqznnnvO+UABj3nwwQd1wQUXKDk5WV27dtW8efO0a9eucz5u/vz5GjJkiOLi4jRy5EgtWbIkBNEC3tKW/rdt2zbdcMMNys7OlmEY+sMf/hCaYAEPaUvf+7//+z9Nnz5daWlpSktL06xZs7R+/foQRQx4R1v637/+9S9NmDBBqampSkxM1JgxY/T888+HJF4SEfC84uJijR49Wn/5y19atP0f//hH5eTk1P4cPnxY6enp+sIXvuBwpID3tLb/ffjhh7r11lt11113adu2bZo/f77Wr1+vr3/96w5HCnhLa/veo48+qvvvv18PPPCAtm3bpp/+9Ke65557tGjRIocjBbxl5cqVuueee7R27Vq9++67qqio0OWXX67i4uJmH7N69Wrdcsstuuuuu7R582bNmzdP8+bN09atW0MYORD52tL/SkpK1K9fPz300EPq3r17CKMFvKMtfW/FihW65ZZbtHz5cq1Zs0a9e/fW5ZdfrqNHj4YwciDytaX/paen64c//KHWrFmjTz75RHfccYfuuOMOvf32247Ha5imaTr+KkCYMAxDr732mubNm9fixyxYsEDXX3+99u/fr6ysLOeCAzyuJf3vN7/5jR599FHt3bu3tuzPf/6zfvWrX+nIkSMhiLJjKCoqql0jIikpye1w4LCW9L2pU6dq2rRp+vWvf11b9p3vfEfr1q3TqlWrQhAl4E2ff/65unbtqpUrV+qiiy5qcpubb75ZxcXFeuONN2rLJk+erDFjxuixxx4LVahAyITqc0hL+l992dnZuu+++3Tfffc5FhPQEbS270lSVVWV0tLS9Mgjj+jWW291OELAu9rS/yRp3Lhxuvrqq/Xzn//cwegYEQGc05NPPqlZs2aRhABCYMqUKTp8+LCWLFki0zR14sQJvfLKK7rqqqvcDg3wtLKyMsXFxdnK4uPjtX79elVUVLgUFRD58vPzJVlXnjVnzZo1mjVrlq1s9uzZWrNmjaOxAV7Xkv4HoP21pe+VlJSooqKC/gqcp9b2P9M0tWzZMu3atatViYu2IhEBnMWxY8f05ptv6mtf+5rboQAdwrRp0/TCCy/o5ptvVkxMjLp3766UlJQWTy8DoG1mz56tJ554Qps2bZJpmtq4caOeeOIJVVRU6NSpU26HB0SkYDCo++67T9OmTdOIESOa3e748ePq1q2braxbt246fvy40yECntXS/gegfbW17/3gBz9QZmZmo8Q8gJZrTf/Lz89XUlKSYmJidPXVV+vPf/6zLrvsMsdjjHL8FYAI9uyzzyo1NbVVUzkBaLvt27fr3nvv1Y9//GPNnj1bOTk5+t73vqdvfvObevLJJ90OD/CsH/3oRzp+/LgmT54s0zTVrVs33XbbbXr44Yfl83HdCtAW99xzj7Zu3cr0ZoAL6H+AO9rS9x566CG99NJLWrFiRaMRugBarjX9Lzk5WVu2bFFRUZGWLVumb3/72+rXr59mzpzpaIwkIoBmmKapp556Sl/96lcVExPjdjhAh/Dggw9q2rRp+t73vidJGjVqlBITEzV9+nT97//+r3r06OFyhIA3xcfH66mnntLf/vY3nThxQj169NDjjz+u5ORkdenSxe3wgIjzrW99S2+88Ybef/999erV66zbdu/eXSdOnLCVnThxgoVzgTZqTf8D0H7a0vd+85vf6KGHHtLSpUs1atQohyMEvKu1/c/n82nAgAGSpDFjxmjHjh168MEHHU9EcIkb0IyVK1dqz549uuuuu9wOBegwSkpKGl197ff7JVnJQQDOio6OVq9eveT3+/XSSy/pmmuuYUQE0Aqmaepb3/qWXnvtNb333nvq27fvOR8zZcoULVu2zFb27rvvasqUKU6FCXhSW/ofgPPX1r738MMP6+c//7neeustTZgwweEoAW9qr2NfMBhUWVlZO0fXGCMi4HlFRUXas2dP7e39+/dry5YtSk9PV58+fXT//ffr6NGjeu6552yPe/LJJzVp0iTmFAXOQ2v735w5c/T1r39djz76aO3UTPfdd58mTpyozMxMt94GEHFa2/c+++wzrV+/XpMmTVJeXp5+97vfaevWrXr22WfdegtARLrnnnv0j3/8QwsXLlRycnLtOg8pKSmKj4+XJN16663q2bOnHnzwQUnSvffeqxkzZui3v/2trr76ar300kvauHGjHn/8cdfeBxCJ2tL/ysvLtX379tp/Hz16VFu2bFFSUlLtlaIAzq4tfe9Xv/qVfvzjH+sf//iHsrOzax+TlJSkpKQkd94IEIHa0v8efPBBTZgwQf3791dZWZmWLFmi559/Xo8++qjzAZuAxy1fvtyU1OjntttuM03TNG+77TZzxowZtsecOXPGjI+PNx9//PHQBwx4SFv635/+9Cdz2LBhZnx8vNmjRw/zy1/+snnkyJHQB+9hhYWFZkFBgVlYWOh2KHBIa/ve9u3bzTFjxpjx8fFmp06dzLlz55o7d+50J3gggjXV7ySZTz/9dO02M2bMqO2LNV5++WVz0KBBZkxMjDl8+HBz8eLFoQ0cCCGnPoe0pf/t37+/ycc0/HwKoHlt6XtZWVlNPuYnP/lJyOMHIllb+t8Pf/hDc8CAAWZcXJyZlpZmTpkyxXzppZdCEq9RHTQAAOggioqKZJqmDMPgiiMAABBSfA4BAKBjYtJfAAAAAAAAAADgGBIRAAAAAAAAAADAMSQiAAAAAAAAAACAY0hEAAAAAAAAAAAAx5CIAAAAAAAAAAAAjiERAQAAAAAAAAAAHEMiApBUVlamBx54QGVlZW6HAnQo9D3APfQ/wB30PcA99D/AHfQ9wD3h1P8M0zRNt4MA3FZQUKCUlBTl5+erU6dObocDdBj0PXcUFRXJNE0ZhqGkpCS3w4FL6H+AO+h76Ojc/BxC/wPcQd8D3BNO/Y8REQAAAAAAAAAAwDEkIgAAAAAAAAAAgGOiWrKRaZoqLCx0OhbANQUFBbbfAEKDvueO+lMiBINBt8OBS+h/gDvoe+jo3PwcQv8D3EHfA9wTqv6XnJwswzDOuk2L1oiomUsKAAAAAAAAAACgRkvWoGhRIoIREQAAeAeLVQMAALfwOQQAAO9pyYiIFk3NZBiG66tqAwCA9uHz+TgBAAAAXMHnEAAAOiYWqwYAAAAAAAAAAI4hEQEAAAAAAAAAABxDIgIAAAAAAAAAADiGRAQAAAAAAAAAAHAMiQgAAAAAAAAAAOCYKLcDAAAAAAAAHYNhGLbfAACgYyARAQAAAAAAQiIxMdHtEAAAgAuYmgkAAAAAAAAAADiGRAQAAAAAAAAAAHAMiQgAAAAAAAAAAOAYEhEAAAAAAAAAAMAxJCIAAAAAAAAAAIBjSEQAAAAAAAAAAADHkIjwMtO0fsJRuMYVrthf4Yc6AQAAHYTJ5x4gJOhriBSmadJe0e5oV95nmNSw91QFpc37pEC55PNJPsPtiOyqgtZJXL9PMsIstnBjSKpkf4UVn6QK6gSRrWhQN5nRflUEpb2FMW6HAwAIc9E+Q0HTOkEQdDsYwMNifIaqTFOmKfoawppPkt9nyDSlSk4rop3UtCu/z9DwblHycb7Fc0hEeE1OrnTwc6myyrq9cY/043+4G1ONe+dIl4ySov3W7aeWSq+sdjemcOX3Sd+/Xpo6xPq3JP3+dendLa6G1aH5fdL3r5OmDq2rkz8ukt7e7G5cQBsUbXxYZo80HS8KatZzxW6HAwAIYxf3i9W3L+wkSaoKmlp3uFy//qBAlZwlBdrVTSMT9NWxiZKkiipTy/YG9Je1RS5HBTTtfy9L0ege1gVN+YGgnt9crLd3B1yOCpHup7NSNC7Tald+Q+qW7FNGot/lqNCemJrJawyjLgkhSRMGSL0z3IunvpiouiSEJM25oO6ELuyqglJirH3/zJvkXjyw6iS+QZ3MpU4AAIC3XTs0vvbffp+hxBiDJATQzqJ80tWD42pvR/sNRYfbzAZAtexUf20SQpJS4nwqr+IaZ5yf3in+2iSEJFWZYkSEB3EW2Gu6pEhRDbKF1050J5aGFq6z3+6SYl3xj6YtXG+/3bebNCrblVBQ7fUGdZLdVRrd151YAAAAHDasa5QGdI62lS3aWepSNIB3TcuKVXqC/Xv86/Q1hKk59RLUkpRXGtQHB8pcigZeMWeIvV35DSk1nkSE15CI8Bq/T+qeZi+7dLSUFNf09qH02TFp+2F7GVeUN2/jbunoaXsZ+8tdm/ZIh0/ZyxipAgAAPOraoQm22zmFVdpwpNylaADvmtvgxO7W4+Xal1vpUjRA81LiDM3sZz+/tGRXKSPlcF6SYw1d0t/erjon+hgR4UEkIryoR5q1yHGNuGjpinGuhWPTcFTEsN7SoEx3Ygl3phrvr0mDGieaEDrN1UlmuivhAAAAOKVrok+Te8fYyhbtKFWQ2TeAdjW0S5QGZthHHi3cwWgIhKcrBsYrxl93wqmiytRbn9FecX5mD4xTbJQ96dA5gVPWXkStelFstJTRyV4WLusxrN4pfZ5vL+Mq/+Yt/Vgqqrfgk8+Qrr3AvXggLftEKmzwQWsOdQIAALzl6iHx8tebo76kPKile1mIFGhv1zYYDXGisErrGXmEMBTlk64abL9q/f39ZToTIEONtvMb0lWD7X8HU+MMRfsZDeFFYXBmGo7I7Gy/HS7rMVQFpUUb7GXTh0mdk92JJ9wFKqS3N9vLLh8rxcc0vT2cV9ZMnSTGuhMPAABAO4uLki4faD/Z9O6egEorONkEtKcuiT5N6WP/HrFoJyOPEJ4uZC0TOGBKVqy6JNrbVUYip6u9ipr1quR466e+cBl58NZH1gn2GlF+6eoJ7sUT7hattxI4NRJirRPfcE/DOomPoU4AAIBnXNI/TkkxdV8Vg6apNzjZBLS7qwc3GHlUEdS7exh5hPDUcPTOp6xlgnbQcI2chGhDCTGcrvYqatbLGs5bHy7rMRQFpPc+tpddOV6KiXInnnB3Ml9au8tedu0F1jRNcMfnBdY0Y/XNoU4AAEDkMyTNGWI/KbD+SLmOF7ESKdCeYqOsedHrW7a3TCWMPEIYamotk9dZywTnaVBGlIZ0sbcrRkN4G7XrZZ07NT65Hy6jIhaut99OSZAuHulOLJFgQYMFknukSxMHuhMLLA0Xre6eZi1cDQAAEMHG9YxRrxT7d4jXt3OyCWhvl/SLU1Ks/ZTMIk7sIkxdOzTBdpu1TNAeGo6yifZJKXFc4OllJCK8zGdYJ6zrC5f1GA6fkjbttZeFS5IkHG07JO3JsZfNnexOLLBsPyztPmYvm0edAACAyNZwioT9uZX69ERFM1sDaAtDjU/AbThSppzCKncCAs7CWsvEvk4la5ngfKXH+zQty75GTudEnwyDRISXkYjwuu6p9uliwmk9hoZXlGd3lUb3dSeWSNBwVMTobKlvN1dCQbWGdTIyS+rX3Z1YAAAAzlOfFL/GZtpPNrEQKdD+xmZGNxp5tJDREAhTrGUCJ1w9OE5R9dqVISk9gdPUXkcNe110lNQ11V4WLusxbNpjjYyobx6jIpr1/jYpt8heNneiO7HA8sE2KbfQXkadAACACDWnwRXa+YGgVu7jZBPQ3hpOc3Mwr1If5zDyCOEnrqm1TPYEWMsE5yXWL10xyP6ZIy3BZ0tMwJtIRHQEDRetDpf1GExJrzdYK2LSoMbxwlJZJS3ZaC+7eKRVn3BHZVBa3KBOZo6QUhPdiQcAAKCNkmMNXdzPfrLpzc9KVcEa1UC76pXi1/iejDxCZLikv30tk6BpatFOEtQ4PzP6xalTnP2UdAajIToEarkjSIhtfGI0XK7aXvaxVNjgQ9ecC9yJJRIs2SRVVNbdjo6SrhrvXjyw6qScOgEAAJFt9sA4xUbVXYlYUWXqzV2cbALa25wh9quACwJBrWDkEcKQocbtdeORctYywXm7tkG7Soo1FBfNaIiOgERER9FwlEF2t/BYjyFQIb292V52+VgpMbbp7Tu6M8XSiq32sqsnWGt/wB35JdLyT+1l1AkAAIggfsOaA7y+VQfKlFvKcAigPSXFGLq0v33k0Vuflaqc87oIQ+MyY1jLBO1udI9oZaXZ21UXRkN0GNR0R5GWJMXbh3+GzXoMb2yQqup9yYmPsZIRaFrDRb7Tk6WLhrkTCywN6yQtSZox3J1YAAAAWmlqVqwyEu0XUTBVDND+Go48qgyaWszII4Spa4fZE9QH8ir1yXHWMsH5mdtgPapYvzUiAh0DiYiOwjCkjE72sgsGNk5OuOFkvrTjiL3sIk7iNmvficaLfF80wp1YYDlwUjr0ub1sBnUCAAAiw0XZ9tHIxwqqtOd0ZTNbA2ir6X3tfW3biQpGHiEsJcUYGpdpP1/0wYEyl6KBV8RHG43WyEmJ98kwSER0FCQiOgrTlE4X2ss+3i+VlrsTT31pSdLgnvayNbvciSUS9M6wfupbs9OdWGDp2Vnq08VeRp0AAIAIsfaw/eRSZie/+qQyzSTQ3tYesn//Hto1WilxnIBD+CkqN7X1uL29Tu4dBheyIqKVVpiNRtUUBIIyTdOliBBqJCI6ivwSqaRB9rrhdDJuuXqCFF3vi05ZhfTWJvfiCXfXNlhovKBEWvFp09siNBou/l5YKr33iTuxAAAAtNL7+8t0psFV2dc2mDoBwPl767NSVVTVnXCL8Ru6YiB9DeGp4RR9AzOiNbRLVDNbAy3zeoN1RgKVUnE5iYiOgkRER3Es13776Glpw253Yqkv2i9dNd5etvxTqYA5aZuUFCddOspe9uZHUhlD512TFCfNGm0vo04AAEAEqQhKb35m//w9s2+cOjFnM9CuzgRMvb/ffoHgVYPjFMWZGYShdYfLdaLQvpI6SWqcr41HynWswN6uThUzRV1HweGuIygtl3IbTMu0cL0UDgnHmSOk1ER7WbiM1AhHV4yT4uoNh6ysshb7hnsuH2uvk6ogdQIAACLOkl32K7VjowzN5kptoN01vMo8PcGvC7Nim9kacE/QlBY1aK9T+sSqSyKnEtF2pqRFO0tsZQVlpsoqw+EkJZzGX4+OIKfBaIjigLTsY3diaWjuJPvtzfukg583vW1H5zOkay6wl63a0XjtD4SOz5CubVAnH+6QThW4Ew8AAEAbnQmYjRYivXpInPwMigDa1b7cykZz73OVOcLV0j0BlVbUnSD2+wxdPZj2ivOzbE+ZisvtoyBOlzAqoiMgEeF1lVXSiTP2src3h8ci1SOzpH7d7WWMhmje1CFS1xR7GfvLXVMGS11T7WULqBMAABCZGs7b3DnBr2lcqQ20u4U7mHsfkaG4wtTSvQFb2eyBcYqlueI8lFaaenePvV3llgRVFWRUhNeRiPC6E2esqWJqVAWlRWEybcy8BqMhwmXdinA1b7L99o4j0q6j7sQCy9wGdbLrqLTziDuxAAAAnKe9uZXaeoIrtQGnrT/S1Nz7CS5FA5zdogaJs6RYny7pF+dSNPCKN3aW2hIPQVPKK2VUhNeRiPAy02w8LdPaXY1HSLihe6o0abC97PUwWbciHA3MlIb1tpcxGsJdA3pII/rYy6gTAAAQ4RqecBrcJVqDM7j0FWhPTc+9H8Pc+whLOYVV2nDEPnXfnCHxYuY+nI8TRUGtP2K/+OFUcVCmyYlBL+Mo52W5RVKgwl4WLidK50y05tevURyQlobJuhXhaO5E++1TBdZaBHBPU3Wyars7sQAAALSTtYfLdaKo4ZXajIoA2tu7ewIqqai7+pe59xHOGk4n1js1SmMzo12KBl7RcErI8iqpsIxEhJeRiPCyY6ftt/fkSFsPuRNLffEx0uyx9rJwWbciHKUnSRcNt5ct2mCfcguhlZYkXTTCXvbGRqmSOgEAAJEtaFrTJdQ3LStWGQl8dQTaU0mFqWV7mHsfkeHjnAodzKu0lTGdGM7X1hMV2ptrv4D6VDHnVbyMT5NeVRyQ8kvsZeEyGuKyMVJCvUXvwmndinB09QQpyl93O1AhvfWRe/FAumq8FF2vTsoqpLc2uRcPAABAO3p3d0ClFXVXJPp9hq7iSm2g3S3aGVCw3jQkzL2PcPZ6gyT1+J4x6pXib2ZroGUaTglZVG7aPoPAW0hEeNWxBmtD5BVJK7e5E0t9PkO6tsGUNuGybkU4iomyTnrX994nUmFp09vDedF+KzlU3/JPpQLqBAAAeENxhalle+1Xal8xKE6xnG8C2lVOYZU2Hmm8QDxz7yMcrdgXUEHAfrX6nCEkqXF+3t9f1miR6lPFVc1sjUhHIsKLKiqlk/n2ssUbpcow6MgXDJQy0+1l4TJSIxzNHCGlJNrL2F/umjlCSqVOAACAtzW8QjE51qeLuVIbaHcN597vlRKlcZkxLkUDNK+8SnrrM3t7vbR/nJJiSJ2h7SqC0psN2tWZUlOVVYyK8CISEV6UkyfVX2W+olJaEibTxsydZL+9N0zWrQhXDffXR3ulw6fciQWWaxvUyeZ90sHP3YkFAADAIccKq7ThSJmtbA6LVgPt7pPjFTrQYO59+hrC1eJdAVUG6843xUYZunwgSWqcnzd3laqiXuLBlHS6hLUivIhEhNcETSsRUd+KrdKZYnfiqS+7qzSmr71sAVeSN2t0ttS3m72M/eWukVlS/+72MkZDAAAAj3q9wZXafVKjNLZHtEvRAN7VcATS+J4x6s3c+whDuaVBfXjQnqS+Zki8/AyKwHk4EzD1/gF7uzpdErStoQNvIBHhNafyrREQ9S1c704sDTW8uj9c1q0IVw3315FT0qY97sQCy7wGdXL0tLRhtzuxAAAAOGxLToUOnbF/t7iWK7WBdrdiP3PvI3I0TFJ3SfRrSp9Yl6KBVzRMyFYGpfwAiQivIRHhNeWVUn6xdKh6qphPD0r7jrsbU42kOGuditxC6/eSTeGxbkU48vukuBj7/np9vTU+De6gTuBBXLgEADiXmhNOlUFTuSVV8vukKL5FAu2q/tz7NX0tnsFHCFOfnarUzs8rJFntNaegSmnxfLPA+dmbW6mtJ8olWe0qr6RKpeVMz+Q1hmkyzsUTTFM6cFJas0vask8qKZOumywdOR0e0zLVqKiUPj4gXTdJyjlDIuJcqoJWfc6dJB3LtabegruqqqQt+6kTRLSiL0yWmRyn4kBQf9/BlwYAQPN8hnRhdpz25VbojR2lGpgRpcQYMhFAe4v1S1Oz4rTj8wot2VmqwV3oawhf3ZL8GpQRpe0nKvTW7lIN6xpNe8V565ro05Au0dp1qkIni4P6ythEdUtimjoviXI7AJynqqC047D04U5p9zHrm8KAHlKvztZJ/u6p1k+4yCuy4i2vlAZ0P/f2HV1ekbTNbyVwBmW6HQ2k6jo5RJ0gohlVQVUWlMqIjtG4ngyjBgCcXVF5UDF+Q+kJPo3sHqPUeE42AU7ILwsqIdpQRiJ9DeGvuMJUSrxP3ZL8tFe0m6IKU4kxPvlYrNqTSEREqrIK60r5VTulo6ekxDhpdF+pa4pkhPHVrYZPkmH9NjhInRP7K/xQJ/CAxNW7tX/jYR278WIZXeLcDgcAEAEMQ9ZHICO8v24AkY6+hkhCe4UTaEveRSIi0uSXSBt3W1MwnS6QuqRIU4dIqYluR9YyRr0fzuGeG/sr/FAn8ALD3pQBADgXQxw7gFCgryGS0F7hBNqSd5GIiBQnzkhrd0kb90jFAWvqpdGjpIQIm1LDMOw/ODv2V/ihTuAFhiHDsGbz89GMAQAt4KtOYnPsAJxFX0Mkob3CCbQl7yIREc5MU9p/whr98OkBaz2Ift2lft2k6AitupoTt5zEbRn2V/ihTuAF9dourRgA0FocO4DQoK8hktBeAZxLhJ7N9riqoLT9sPThdmnfCSnaLw3pJfXpEvlpQZ+sE2A+Rf57CQX2V/ihTuAFPsmQIcOwfgAAOJeaYwbHDsBZ9DVEEtornEBb8i4SEeEkUC5t2S+t3ikdPS2lJEjj+0vdUr1z5bVRMzE5V5O3CPsr/FAn8ILqdkwzBgC0FINCgdCgryGS0F7hBNqSd5GICAf5JdKG3dK6XVJukZV4mDEichagbg2jegkjjlItw/4KP9QJvMAwatdbZ811AEBL+CSOHUAI0NcQSWivcAJtybtIRLjpeJ607jNp016ppEzK7iJdMDDyFqBuDVYyah32V/ihTuAFPoOBPQCAVjEM+w8AZ9DXEElor3ACbcm7SESEmmla6z6s3SltOySZkgb0kPr3sNaC8DqOUq3D/go/1Am8gKmZAACtxPQbQGjQ1xBJaK9wAm3Ju0hEhEpV0Eo8rN4h7T8pxUZLI7KqF6DuSIOOjLrf/GVpAfZX+KFO4AXVIyJU16IBADgbo8EPAGfQ1xBJaK9wAm3Ju0hEOC1QLm3eJ63ZKeXkSenJ0pTB3lqAujV81SdvmdamZdhf4Yc6gRdUT81EMwYAtBQfgYDQoK8hktBe4QTakneRiHDKmWJrAer1n0lnSqTMNGnWaG8uQN0apMtbh/0VfqgTeIFR9w+DhgwAaJG6UaEcOwAn0dcQSWivcAJtyatIRLS343nS2l3WKIiyCqlvd+nCYVK8hxegbg3DJ2tKG1/1v3FW7K/wQ53ACwxfzTIRHXJwHgCg9WzXYnDsABxDX0Mkob3CCTQl7yIR0R5MU9p73EpA7DhsjSEalGktQh3NLrapPocrX/UPzo79FX6oE3hBddtlUTkAQEsZ1WeaOHYAzqKvIZLQXuEE2pJ3cZb8fFRWVS9AvVM69LmUECuN7SdldbQFqFuh/iW4/GU5N/ZX+KFO4AUGi1UDAFqH2SmB0KCvIZLQXuEE2pJ3kYhoi9JyafNeac0u6eQZKSNFmj6s4y5A3RpG9f8Yt9cy7K/wQ53AC1hUDgDQSr7qjz8cOwBn0dcQSWivcAJtybtIRLRGzQLUG3ZLBaVS787SleNZgLo16l9Nzl+Wc2N/hR/qBF5QnURjYA8AoKWYfgMIDfoaIgntFU6gLXkXiYiWyMmV1n4mbdknVVRZaz9cPNKaigmtYxj2H5wd+yv8UCfwAsOgGQMAWoXZKYHQoK8hktBe4QTakneRiGiOaUp7cqwFqHcdlaL80tDe0kAWoD4vNX9NOEq1DPsr/FAn8IKaERFi/k0AQMsY9X5z7ACcQ19DJKG9wgm0Je/ijHpDlVXS1kPSmp3S4VNSUrw0cRALULcXriZvHfZX+KFO4AWGIUOGfIb1AwDAuVgffYzq2Sk5dgBOoa8hktBe4QSakneRiKhRfwHq0wVStzTp0lEsQN3eWMmoddhf4Yc6gRf4DNt8rgAAnAvTbwChQV9DJKG9wgm0Je8iEZFXVLcAdUmZ1KerNGUIC1A7hZWMWof9FX6oE3gBUzMBAFqJ6TeA0KCvIZLQXuEE2pJ3ddxExLFca/2HTw5IQVMalCkN7y3FswC1o7iavHXYX+GHOoEX+AyuXAIAtAqzUwKhQV9DJKG9wgm0Je/qWImImgWo1+yUdudIMVHSqGxpcE9rMWo4j6NU67C/wg91Ai8wrESET5LpdiwAgIjgk2qPHaycBziHvoZIQnuFE2hL3tUxEhH1F6A+elpKSZQuHMYC1G6oOXHLSdyWYX+FH+oEXlA9mSvNGADQUnwEAkKDvoZIQnuFE2hL3uXtRERJmbR5nzUFU16R1CNdmj1W6p7mdmQdm9HgN86O/RV+qBN4QN3UTDRkAMC5GYZRe9zg2AE4h76GSEJ7hRNoS97l3UREfon06JtScUDq102aOYIFqMOBT9aZL5+YX78l2F/hhzqBF1SPoa5JRgAAcC5c9QqEBn0NkYT2CifQlrzLu4mIkoD1c+U4RkCEE45SrcP+Cj/UCbyANSIAAK1k1PthclvAOfQ1RBLaK5zAmRbv8m4iokZ0FCcLwwkL/bYO+yv8UCfwgupEhGjGAIAWqp3STxw7ACfR1xBJaK9wAk3JuzyeiDCsqVOYPiV8+KqPUNRLy7C/wg91Ai+obrs1Vy8BAHAu9ZfI4tgBOIe+hkhCe4UTaEve5e1ERN1KnG5HghpG9eGJemkZ9lf4oU7gBbWLytGMAQAtU/+KV44dgHPoa4gktFc4gbbkXd5OREjVi8rSgsMGV5O3Dvsr/FAn8AKfIUOGfHxjAAC0kM8wZBjWscPHsQNwDH0NkYT2CifQlryrAyQixEmWsGLYryjHObC/wg91Ai9gaiYAQOsw/QYQGvQ1RBLaK5xAW/IubycimHci/NTUBfXSMuyv8EOdwAuYmgkA0EpMvwGEBn0NkYT2CifQlrzL24kIielTwo3PqJsui3o5N/ZX+KFO4AU1i1XzhQEA0EIskwWEBn0NkYT2CifQlrzL+4kI/hqGF9LlrcP+Cj/UCbzAMGqHT9OKAQAtYTT4AeAM+hoiCe0VTqAteZe3ExGcLAw/pMtbh/0VfqgTeEG9qZkY2AMAaIn612Jw7ACcQ19DJKG9wgmcavEubycixPQpYccn6yjlE/XSEuyv8EOdwAt81q+avBoAAOfCtRhAaNDXEElor3ACbcm7PJ6IEOPDwg3j9lqH/RV+qBN4gUETBgC0Dh+BgNCgryGS0F7hBNqSd3k7EWFU/y/cUmnzP5S+90zd7ZgoqWe6NH249B/XSF06uRaa45xOl9fs29d/KI3Kbnz/zb+W8oqkd35qL68KSlO+L53Ml57+T+nike0fW1t47fKC55dLP/qHNLqvtPB/mt7m6GnpL0ukldukz/Ol5HhpXH/pG7OlCQNCG29TvFYnaJtzteXsr9tvx8dIPTtLcydKX7tMio8NTZzNqR5D7eaIiIXbS/WTpQWSpKdvTNPYzBjb/aZp6oqnT+lEUVDTs2P052vTJElj/nRCN4+K1/0zPXysBIAwVPOxx82PQDXHjhduTtfwbtFtfp7SClPPbCrWhF4xuqBXzLkfAIRQOPS1Gv/8pEQPrijUiG5R+vvNnRvdP+ZPJ2y3E6MNDekSpdvGJ+qivi5/3kVIhEN7ba9jA8KH23/74BxvJyKk8Dh6N1QTz7fnSr0zpLIKaeMe6e8rpOWfSu/+1P2TVI4xnE0QnesoaDTYrsaanVYSoleGtHCddMmo9o+tTRzeX6G2YJ21jz/eLx08KWV3s9+/Ybd0x5+sf998oTQw00pGvLJa+sLD0k++KN1xaejjtvFYnaBtztWWJWn6MOn6Kda/S8qk9bul3y6UdhyRHr07tPE2YlTPMGbIdCkTUT07lGL90pu7AhqfaT/ubTharhNFQcX4rWh99eJseBsA4DxfvWOHW3+DfbW/zy+G8kpTf1tfLEPSpF5e/d6FSBUOfa3Gm7sCyuzk19YTlTpypkp9UhufQprSO0ZzhsbLNKWcwiq9/GmJ7l10Rn+dm6ZpWfQvrwuH9tpexwaED+rRu7yfiAjHNSJqwrlklDQ62/r3l2dIaUnS/70jvfuxNG+SW9E5y2fUrWLkRL3UTzQ0+fxGXRz1LVgnjcySbpwq/epfUqBcSgiDD01O769QOvS5tGmv9Pg90v3PWQmf/5pbd/+ZYunfH5PioqV/3S9ld627799mS1/5nfSzl6w+4+bICC/VCdrmXG25Rr/u1t+UGrdeLFVWSm9+JJVXWm3dLT6jdlE51z7jVb/u9OxYvbsnoPtndlJUvT615LOAhnWN0pnSoLV5gzjJAwJAaNVc5+PqdV7tdOVtOFzBCzQnLPqapCP5ldqSU6E/XJ2qn71XoCW7Arp7clKj7bLSojRnaHzt7csGxmnu86f0wpZiXZgdBt+p4aiwaK/8Tfcc6tG7vJ2IcP2vYTNqP/nKHtuFQ61ExJFT1lRBjyy2phrKyZO6pkjzJkv/da0UW+/k1eTvSYN7SnfOkn4xX9qbI/XpIn3veumq8SF9Wy3i9Kf+5vZtw23q31daLr31kXTvHGnOROmnL0nvbJGum9z+8bWWl74lLVgnpSRKs0ZLqyZIr62Tvj2v7v4XVlqjUv7wNalvg6vLE2Kt8un3S39YJL3w7ZCGbuOlOkHbnKst19ewjXRJtcqi/e62n5qpmSTXEhE1L3vVkHgt21umtYfKNb16CH9Flal3dwf0jUlJemFzcaMBSHQ/AAi9cPgIZNT7R1MxVFSZ+tv6Ir2/v0yHzlSpKigN7Rqlb01J0sTe1jHmaH6lZj99SpL06LpiPbquWJJ096RE3TMlOQTvAji7cOhrkrRkV0CdYg3N6Beryw7HavGuUv37lMaJiIZx9u8cpbR4Q0cKqvi81gGEQ3s917FBknacrNAfPyzU5pwKBU1pVPdo/efUJI3uUTc934JtJfp/7xbouZvStXR3QIt2lipQIU3JitEDl6YoPcHX9JOj3fG3w7u8nYiQwvOq5Zoe1TC2g59bv9OTpO8/I738oXTNBOkbV0ib91qJiT050lP/YX++/SetK8m/OlO6aZr0z1XSN/8qvfAdacbwULyjlnN8RET1cxYFrLUgGqqsqoujxtItUnGZNQqle6o0dYi0YK10w5T2j6+1vHT1/WtrpavGWVeBXzdZem659Ml+aUw/6/6lH1v3zZ3Y9HvN7ipNHCit3mFNZxbv0ny+XqoTtM252nKN8oq6v0Ol1VMzvfKh9ZgYlw+/YTQiomcnv0b3iNabn5Xqon7WSaJVB8tUVG7qqsFxemFLcZO5ZT6cAkBo1fwtPtv1PiEJQs3HUFxh6tWtpbpqcJxuHBGl4gpT/9paon97LU8v3dJZQ7tGKz3Rpx9f0kk/e69AswbEataAOEnSoIwoji0IC2HR1yQt3lWqywbGKSbK0NWD4/XPT0q19USFRnZvPKq3fpyFZUEVBEz1TvHRpzqAsGiv5zg27DlVoVvn5yopxtCdExIV5ZPmf1KqO17J1bNfSNeommRE9WMfXFGgTrE+3T05SccKqvT8RyX6pb9Av706NRTvBnLvKyqc5/1EhNtH76bUhFMYkHKLrJOqG3ZLv39diouRBmRKP3hO+tJF0m/vtLa941Ipo5P06FvS6p3StKF1z7fvuPTEt6SrJ1i3vzTDunL8F/OlmSNC+tbOqf4RypEREdW/b/5189sM7ml/7X+tsab66ZVh3Z47Sbr/eel0obXP3eT0/gqVjw9YSbRffMV6H5MGSZnp0r/WSmP7W9vsPib17271geYM6yOt2WXNyT+0d0hCb8QrdYK2aUlbrvHiB9ZPfVeMk35zh/ttx6hORMi9D3m+er+vGRKn339YpPJKU3FRhhbvCOiCXjHqnuRvtL2k6nloAQChVP+44dbf4PrHjqZiSI01tPSuLorx1x3dbhoRr6ufPaV/bCnRLy5PUVK0T1cMitPP3ivQ4Iwoza03pQwQDsKhr207UaF9uVX64cw4+SRN6Bmt7kk+Ld5ZqtENEhHlVabyS4O1a0T8cXWhqkxp9sA4Pq91AOHQXs91bPjT6iJVBk29cFNn9a5e52Te0Hhd9ewp/faDQj1/U2fb86TG+fTk9Wkyqr+zmab0980lKi4LKjmWVh0KnGnxLm8nIsL2ZGF1PDc9bC/ulSH99ZvSpj3W7W9eaY/97iutRMTSj6ULh9U9V/dUKwlRs22nBOkL06wRFJ/nS11THXwvrVRz+a1j9VKTwr7VOqnd0AMvWtNe1bx2bpG0Yqv00y/VlV1zgfQ/z0uLNlhTXrnJ8f0VIq+tkbqkWO225r1cO1F6dY217/0+axRLUvzZ32eydcWaigIujvv0SJ2gbVrSlmtcMa7ub0hpWfW6Em9L9/zNSh6Hw9RMYTAiQoZ01eB4PbSyUCv3l+nC7Bit2B/QDy/u1Gg6JtvD6X4AEFLhMP1G/WNHUzFE+Y3aL7hB01RBwJQpaXi3aO04WWF7D7VPyfEEYSYc+tqinaXKSPBpcp+Y6jgMXTk4Tot2BPTfM5Llrzcy/NWtpXp1a2nt7Wif9LUJibpjQgL9qwMIh/Z6tmNDVdDU6oPlmtU/Tn3S6k6Bdkv265ohcZr/aamKy4NKivXVPs/NoxLkq9fGJ/SM0bMflSinsEqd4khEhAJ/O7zL24mIcD1ZWBPPQ7dai5lG+aUunaQBPSSfT3pzkzXlS79u9ti7pUkpCdLR0/X+2suaT9/X4I9hzUn4w6etx4ULx0dEVD/nuH6Np0mRpNREa6RDzXavr5MqqqRR2dKBk3XbjetvjZS467L2j7E1vHD1fVXQmlN/2lDp8Km68vEDpMfeklZtl2aOlJLizp1gKCqzfp8rYeEkL9QJ2qalbblGZro0o96otCvGS+nJVkL03Y+l2WNDF3tDhqGa/9zKRBjVr2tI6pzg15Q+sXpjR6kCFaaqTOmKgfGyR2nUe6z9NgDAeUaD/9yKwfrd/HHgtW0lempTsfbnVqoiWFfeK8Vf7zE1z+PeewGa43ZfqwqaWrIroEm9Y3Q0v64Tje4Ro6c3lWjtoQrbItSX9o/VV8YkqqLK1KcnKvTY+iIFKk35DU7YdgRut9eaGKzfjY8NeaVBlVaa6psW1ei+/unRCpqlOl4Y1MDYumNEZrLftm1KdfKhIGByzAgR9rN3eTwRIdXO5R5OasIZ37/pk+U1Jzf9vuZjb1je8HZz61C4rebkrVP1UvOUZ3vfhurue3WN9fuanze97aHPrbUJ3OL0/gqF97dLJ85Y624sWNv4/lfXSJeMkgZlSp8elCoq7Quy17fjsLXI74Ae7u0PL9QJ2qalbbm+hm3koup1e9btkq4c50iYLVJ9jHBzRETtqaDq7jRnaJz+3zv5OlUS1EXZsUqJ99k2tuX9yAMCQMiFw6DQhseOhhZuL9F/v52vWQNi9bULEtU5wS+fIf1tfZEOn6lbONeo94QcTxBu3O5r6w6X6/PioBbvCmjxrkCj+xftLNX0vnWJiO7Jfk2rTkzM7B+ntHiffvZegSb1idHsgUx95nVut1fp7MeGs43YaHhfzd1+XzMj5zhmhAz72bu8nYgI16uWz/VXrHeGFDSl/SekQT3ryk/mS/klUu8u9sftP9ngeWWtGyFJfbqE1/uvfxGSkyMiznqEqL7v4ElrbY6vXSZNHWrfJBi0FgD/1xrpO/PaP86Wcnp/hcKra6wRP7+6vfF9b2yQlmyUAndKl4+VNuyRXl8v3XRh420PfS6t3WUtwJ4Q2/j+UPFCnaBtWtqWaxZSb+r4U1V9VVlxmbvtp/6XBbfCaPDB//KBcfrxu/naklOhP16TetapmOh+ABB6YTEo9CwnlCTp7d0B9U7x669z6+b2lqQ/ry6sfZxUN5ic4wnCkdt97fWdpeqc4NNPLm28XuI7uwN6d09AZZWm4qKrL2yRPc5bxiTomY+K9YdVRZo9MM7WF+E9brdXK4h6sTSIoXOCT/FRhvbnVTa6b19upXyG1KOT3/a9qOHzcAog9NjP3uXtRIQUnomIGs3FdtkY6X9flv72tvS7u+rKH33T+n35GPvjjudJizdJcy6wbheUSP9cJY3MkrqnWWUVlVbColN8XZkb6l9N7mS9NPf89VPer6y2/v2fc6SenRtv+/cV1jbfvc6xMM8pVPvLKaXl1gnauZOsn4Z6pFnJnrc/km6/1Jo//6cvSZMG20eiBMql/3zcWiXqe9e7fAI3wusEbdOatnzdlLryhm3knc3W7xF9XG/Hhtxd8Ln+B3qfpOQYn342K0VHC6o0q3/cWRendjt2AOiIfKr7++vW3+CGx46G/EbdpIM192/JKdfmYxXK7OSvLUuMsp6psCzY6HkKy4I6WRRU1yQfi5LCFW72tUCFqXd3B3TloDhdPbjxaIbuSX69sTOg9/YGdM0Q6/6G/THGZ+hrExL146UFWranTJcPjAtN8HBFuB8bfD5DF2bHaOmegI7lV6pXinUa9FRxlRbtKNWEnjFKqf5b39zzNFV+sqhKhWWm+qT6Fe3nvEB74+jrXR5PRBjhNzWRVBdPc7GNypZuuUh69j0rqTB1qPTRXunF961FqevPOS5Z09Tc+7i0ZZ/UNcU6gf55vvSXb9Q9/4kz0pTvWc/71286+ObOwVedLneqXs61b+tv9+pqK1nTO6Ppba4cL/3gWenTA9Lovu0eaos4vb+c9vZH1roPV41vOv6JA6WMTlbC54ap0jP3SV/8tXTxD6VbL5YG97Ta7ovvS/tOWOuqTB4c6ndhF+l1grZpbVuWrJFp8z+0/l1aJm3cY7Xlft2kL053t/1Uv7arUzM1ccXRjSMTzrm9dYM8IACEWlhMv1H9uq9sLdUHB8oa3T+5T4ze3h3Q3QvzdHG/OB3Or9Q/tpRoYEaUisvN2sfHxxga2DlKi3cF1Dc9SqlxPg3KiNLgLtF6Z09A338zXw9fmaIbRzR/XAKc4mZfW7YvoKJyU7MGxDX52uN6Rqtzgk+v7yjVnKHViYom4rxxRIL+8GGRHl9fpNmDSER4WSQcG+6blqwPD5br5hdP68tjExVlSC9+XKLyKlP/PSPZ9r2k5ndTUzPVL//NB4V6dVup3v+3LrXJDbQfvut5l7d7S824qbBrwUbdr+Zi+/O/WVeE/+N96yrcbqnSt+dKP7ih8WP6d5cevl360QvSnhwpq4v09L3SrDGNX1Nnec1QcPwo1YJ9a0j6+ID02bGzX11fk4h4+cOm1/IIhXA4qp+P+R9KcdHSxaOajt/vt6Zkmr9KyiuyFgFe9SvpdwusRYFP5EmdEqSJg6RHviFNGRLyt9BIpNcJ2qa1bVmSln9q/UjWRKPdU6VbL5F++AVrwXU3GUZdEsKtLwz1fp+rKzW1Dd0PAEIrHKbfqHnZF7aUNHn/h9/sqpIKU//YUqL39+drYOco/f6aVC3ZFdDaQ+W2uB+6IkUPLC3QL5YXqLxKundqkoZ0jWYKDrjOzb62cHupYqOk6X1jm/7Iaxi6uF+sFm4v1ZlA3ULWDbeNjzF06zgrGbHucJkm93Fxal04KhKODTeOTNDLX+qsh98v1GNrixSUNKZHtH5/TarG9oxp9DwN//43maBoJmmB9sE+9S7DNE3T7SAckZMrPb1Muusy6yS+V434ljS0tzT/B25H0jLH86Snlkp3znJ3iqhIwf4KP9QJvIB2DABopRNFVfrHlmJ9aUyiuiX53Q4H8Cz6GiIJ7RVOoF15F9NuAQAAAAAAAAAAx5CIAAAAAAAAAAAAjiERAQAAAAAAAAAAHOPtxao7gq2PuB0BAAAAAAAAAADNYkQEAAAAAAAAAABwDIkIAAAAAAAAAADgGBIRAAAAAAAAAADAMSQiAAAAAAAAAACAY0hEAAAAAAAAAAAAx3g3EWEYkt+7bw8AAAAAAAAAgEhgmKZpuh0EOpCKSul0odQ5WYqOcjua8Mf+Cj/UCbyAdgwAaKWKKlO5pUGlx/sU7TfcDgfwLPoaIgntFU6gXXkXiQgAAAAAAAAAAOAY5i4CAAAAAAAAAACOIREBAAAAAAAAAAAcQyICAAAAAAAAAAA4hkQEAAAAAAAAAABwDIkIAAAAAAAAAADgGBIRAAAAAAAAAADAMSQiAAAAAAAAAACAY0hEAAAAAAAAAAAAx5CIAAAAAAAAAAAAjiERAQAAAAAAAAAAHEMiAgAAAAAAAAAAOIZEBAAAAAAAAAAAcAyJCAAAAAAAAAAA4BgSEQAAAAAAAAAAwDEkIgAAAAAAAAAAgGNIRAAAAAAAAAAAAMeQiAAAAAAAAAAAAI4hEQEAAAAAAAAAABxDIgIAAAAAAAAAADiGRAQAAAAAAAAAAHAMiQgAAAAAAAAAAOAYEhEAAAAAAAAAAMAxJCIAAAAAAAAAAIBjSEQAAAAAAAAAAADHkIgAAAAAAAAAAACOIREBAAAAAAAAAAAcQyICAAAAAAAAAAA4hkQEAAAAAAAAAABwDIkIAAAAAAAAAADgGBIRAAAAAAAAAADAMSQiAAAAAAAAAACAY0hEAAAAAAAAAAAAx5CIAAAAAAAAAAAAjiERAQAAAAAAAAAAHEMiAgAAAAAAAAAAOIZEBAAAAAAAAAAAcAyJCAAAAAAAAAAA4BgSEQAAAAAAAAAAwDEkIgAAAAAAAAAAgGNIRAAAAAAAAAAAAMeQiAAAAAAAAAAAAI4hEQEAAAAAAAAAABxDIgIAAAAAAAAAADiGRAQAAAAAAAAAAHAMiQgAAAAAAAAAAOAYEhEAAAAAAAAAAMAxJCIAAAAAAAAAAIBjSEQAAAAAAAAAAADHkIgAAAAAAAAAAACOIREBAAAAAAAAAAAcQyICAAAAAAAAAAA4hkQEAAAAAAAAAABwDIkIAAAAAAAAAADgGBIRAAAAAAAAAADAMSQiAAAAAAAAAACAY0hEAAAAAAAAAAAAx5CIAAAAAAAAAAAAjiERAQAAAAAAAAAAHEMiAgAAAAAAAAAAOIZEBAAAAAAAAAAAcAyJCAAAAAAAAAAA4BgSEQAAAAAAAAAAwDEkIgAAAAAAAAAAgGNIRAAAAAAAAAAAAMeQiAAAAAAAAAAAAI4hEQEAAAAAAAAAABxDIgIAAAAAAAAAADiGRAQAAAAAAAAAAHAMiQgAAAAAAAAAAOAYEhEAAAAAAAAAAMAxJCIAAAAAAAAAAIBjSEQAAAAAAAAAAADHkIgAAAAAAAAAAACOIREBAAAAAAAAAAAcQyICAAAAAAAAAAA4hkQEAAAAAAAAAABwDIkIAAAAAAAAAADgGBIRAAAAAAAAAADAMSQiAAAAAAAAAACAY0hEAAAAAAAAAAAAx5CIAAAAAAAAAAAAjiERAQAAAAAAAAAAHEMiAgAAAAAAAAAAOIZEBAAAAAAAAAAAcAyJCAAAAAAAAAAA4BgSEQAAAAAAAAAAwDEkIgAAAAAAAAAAgGNIRAAAAAAAAAAAAMeQiAAAAAAAAAAAAI4hEQEAAAAAAAAAABxDIgIAAAAAAAAAADiGRAQAAAAAAAAAAHAMiQgAAAAAAAAAAOAYEhEAAAAAAAAAAMAxJCIAAAAAAAAAAIBjSEQAAAAAAAAAAADHkIgAAAAAAAAAAACOIREBAAAAAAAAAAAcQyICAAAAAAAAAAA4hkQEAAAAAAAAAABwDIkIAAAAAAAAAADgGBIRAAAAAAAAAADAMSQiAAAAAAAAAACAY0hEAAAAAAAAAAAAx5CIAAAAAAAAAAAAjiERAQAAAAAAAAAAHEMiAgAAAAAAAAAAOIZEBAAAAAAAAAAAcAyJCAAAAAAAAAAA4BgSEQAAAAAAAAAAwDEkIgAAAAAAAAAAgGNIRAAAAAAAAAAAAMeQiAAAAAAAAAAAAI4hEQEAAAAAAAAAABxDIgIAAAAAAAAAADiGRAQAAAAAAAAAAHAMiQgAAAAAAAAAAOAYEhEAAAAAAAAAAMAxJCIAAAAAAAAAAIBjSEQAAAAAAAAAAADHkIgAAAAAAAAAAACO+f+RSqSC2k7NHAAAAABJRU5ErkJggg==" + "image/png": "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" }, "metadata": {}, "output_type": "display_data" } ], - "execution_count": 17 + "execution_count": 19 }, { "metadata": { "ExecuteTime": { - "end_time": "2025-01-10T14:03:21.717660Z", - "start_time": "2025-01-10T14:03:21.298376Z" + "end_time": "2025-01-14T16:33:31.721662Z", + "start_time": "2025-01-14T16:33:31.059791Z" } }, "cell_type": "code", @@ -1213,15 +1419,83 @@ { "data": { "text/plain": [ - "
" + "
" ], - "image/png": "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" + "image/png": "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" }, "metadata": {}, "output_type": "display_data" } ], - "execution_count": 18 + "execution_count": 20 + }, + { + "metadata": {}, + "cell_type": "markdown", + "source": [ + "### Traditional Explanation with Baseline Imputation\n", + "The traditional way to explain any black-box model trained on tabular data is by using imputation strategies for feature removal (excellent [paper by Covert et al.](https://jmlr.csail.mit.edu/papers/volume22/20-1316/20-1316.pdf)).\n", + "During explanations, the model is queried multiple times with different subsets of features removed.\n", + "Removed features are imputed using different strategies, such as the baseline imputation.\n", + "Baseline imputation replaces the removed features with the mean/mode of the training data.\n", + "\n", + "We can natively use the ``shapiq.Explainer`` (specifically ``shapiq.TabularExplainer``) to explain the TabPFN model:" + ], + "id": "b225c897c1181eee" + }, + { + "metadata": { + "ExecuteTime": { + "end_time": "2025-01-14T16:35:57.097208Z", + "start_time": "2025-01-14T16:33:31.724663Z" + } + }, + "cell_type": "code", + "source": [ + "explainer = shapiq.TabularExplainer(\n", + " model, data=x_test[:50], index=\"SV\", max_order=1, imputer=\"baseline\"\n", + ")\n", + "explainer._imputer.verbose = True # see the explanation progress\n", + "\n", + "shapley_values = explainer.explain(x_explain)\n", + "shapley_values.plot_force(feature_names=feature_names)" + ], + "id": "41314e231db2e986", + "outputs": [ + { + "name": "stderr", + "output_type": "stream", + "text": [ + "C:\\1_Workspaces\\1_Phd_Projects\\shapiq\\shapiq\\explainer\\tabular.py:132: UserWarning: You are using a TabPFN model with the ``shapiq.TabularExplainer`` directly. This is not recommended as it uses missing value imputation and not contextualization. Consider using the ``shapiq.TabPFNExplainer`` instead. For more information see the documentation and the example notebooks.\n", + " warn(\n" + ] + }, + { + "data": { + "text/plain": [ + "Evaluating game: 0%| | 0/256 [00:00" + ], + "image/png": "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" + }, + "metadata": {}, + "output_type": "display_data" + } + ], + "execution_count": 21 }, { "metadata": {}, diff --git a/docs/source/notebooks/tabular_notebooks/tabpfn_values.npz b/docs/source/notebooks/tabular_notebooks/tabpfn_values.npz deleted file mode 100644 index 184e0ee67c3636934af20b2037788f4fec43b8d7..0000000000000000000000000000000000000000 GIT binary patch literal 0 HcmV?d00001 literal 2151 zcma)7dpMM78z0AzVV21uIejU%Fb$PM&cuh+4yv)HayHo+Gc;t-5KFe0tV0{pki$+Y zv^C1cfG&&{@u@g!(7G0!2ke2 zN-(tnmNJi6!QueGRl!yU00vz4Lq`VT3^8HRA^_0#TY(!u=pnSbw-!9JgH+mmR8qi- zX`QVOP>7d3NTYDjC9@91N=oz=Nn|Lb{#+W{A9 zw?{WZCsV3Sy~XQsu8Wfmh}alEYTXz2Y$lF5jXS@-jArN6`%f#3XIpk9P>*^t8q5-0 z-%y7Je57UF#a>6fU+jWnCVt z6~K@_!$G7e&9Lqq2nJv%OeK$79K5e1>qmY%J>j8upFkn9Uyiigx|_&3YRq>$ z!G@40X=+q!lbo0a`19rYfpA-z$>hsAl3BV{&$Bfi5W+a)kGL5a-c(9_o?cxh8kaXc z8B_=1@>CqS9eSUa6+X{Go@);4Qx5S@PQjJ8+N{PFbs+jV8)@*5TeH!<*1XzxYxjeD zp(g`0Wu4bw6y)jQR4Fq?wxty+T}+Q);N?ZLPWyrK?SvIaEXejLD%*+9*Y0A~Ki+65rj4MJ_5H z#P5k8J>9EDbw7?Wq}Q#a^dQ=vnF za)i>GF`r_qkZPZq_9ldjvN6ZaCb6>e~*TXKmDk)O9-@Q)$jEIQ^-n^UZzz}p z=ctPYg@&r*+_MQOmZnHYb&6_npYgqG6ck=J8-QJ@R1Zj}5ngUi>QVm~d&N^92221U z9_zz?OFtQQScR;XC0^q~%aBAw+rAIQb5K*!csFyif@9q5H#PVdp3Fvb{HG6Ah*K3a zn9}d1bw_g)0OKB<#ksO&L#AA`7{q~F`QYeyef#=H zcJv>ZRr!|rgDt!TUu^j%xh6#en}wavz}vfN<6c-xHV<{wFx~F?Ab0cQ+n*J%n^qbc zZ*1ZkQk#@hH>{=P?4w$NYqMS;j==}sB_)lZIO4HX}P;AGkx1I|Q z@~Bbm>qDF-%t}LwM;>!7lx!B_#$N=kblIvWozg%qQTi{w;FWYdI8qjacS%EKES;t&| z+Wu-TlaRpse#Y_%w_Y;?M|CmYG{%ZqoL;%N0NPSq8;FV6-SZo!&5re0bnDpAuza(j zE$WSIuIp;v7%g4|wxdFIIxQbYiu|WS$q6b{0M-v35)p#M{9lz4dI_!6EBJt3gFd4&cM?7gdS5__+-rqi*yN+wBz*b@ypyPgs zN(#NI67qmEO@E4czzG>|UjMYfqxeJy@Tv)_l=B2UW`Z)#XT4u}RMwlGH(xV?Q%`!z zqAA-P`bF{IDHi(Gs4ML0$*BbY;!$5Y$Johlkj{2zYQ;)ibR2Z3B^Go86u6#!v5)TA>D^-S`y`ZgP0#ZzR7cQ$-Tw5j^(s0vee;QOzku2_vNxQHD|!#k7`+Y3WsM zNo~{C5+(Mg6dkdYP)pQaTO^8`xX-<}dS{+z-v2rO|9t=VzW;ZgbIyMb${HvD1polT zTvGyQ9jvcr~=mE>19TG zhb3A>a@3-j!V;Gb{jd@gGgh~Ve!D$YPoOXfU0e59>`452f7S6bG76pe7W@FNyx!3_ z(H=ZU=$HNOyt%DC?jkM}bHVntcgB6r9QS(0uz zG+Tlz-}Qy)+%Qy{iQJGgD{VUc;msEH5hL?h$hSS^9(mm?D)s$qmLy$p-CDQ&YUy|+ z{cMSkf0@L)2-Rltw=WOaG5#PmRd!k_Zs?maJ&PY{OP;1&cQmV}7Z9IL_5i!=&g`HR zr!zgM1ti-M8$l00jOCgE@xtfXRa=hH6>Ajxq}AuHQubsyQ=|CP)Mj<`p{{Z=}Y?90F zX0nnN-cNWTwy`v2dghY(PV982m)QsVMgswOPJ7B&r%lEu2)} z>RD0SA@xo4(&sNVnZJCNvikGVH%G}+_ZeiSKy`mTZT0tfqHpb^Xf1G5Z>g5u8|py= z;gPD=;CUyiUIFkRRwsFh-am$%y|dctB1xZ91vjqkc^fMvOI~ zkm?&oZk8b#W_0%_&3Pt7knvoquznc{g9yYIEsh=Kycmb4(bESym$57b3X-_m` z&f*NUz{Y*huiRAem;#f}*Bm|=ASz%7b93epg8Ap1v{+KESL~04_T6O#FJh-O>V{%- zBDU1mYZCP6^RM0HpO_YRU0(Y!+p5h^cgEb$D^OPU6YNEZ4BBGn{ZF5s44O0&Xxj>o zguswA{kIK*`K<}<5r~}nG*X6fb%Q6CT=fdLY(RoXR`Jc2d|ZskrCYlC8H5vSQ{CdV zX$^&d-OG}O=UF?(UBsDNPAXkF&!kN?9No zzX&;jxkRoRFDD}(b0)@L;2(2vBMq#Dx$PQsi05`iluV6lR2ZI7GqSThx%RZAD=l}n za)=-w_hPvlW_mPT@`nUN^$Nd8_QcqS7>~kXcQLDoF>#+Wr_NUai_WT8sJtrT0F@#B z<9>#j;9@}+l-5xEsbqdvj1{WN9v=D)`1U|301TNx0oA13;BcupXLz5c05CQ_{Z@yh zj6{kkGYI+yDs7@sx_tOe#ksW~Z_x$3pw3d{fD3L)(z+1IxPhNVLRugjs$bAdn0!N@eq*-j zk<&7|6(8BEs_?K_@4Q`>NkJPxzBy?bKiS@&wJkUJNNuhc0eI0000o> zO;v#IzN&NnLIA*3-c}3%1%zV)!>|MdJ|t2A0NH)Za|7@__?G|6mgApc(Dp{{lLS;p zDBL1C#d z2c%oYax`ODqSEL0|F9BrZM=RF{d#Mvfk0sqxz{EaOZI0m@`iA9bVR(tB&|r(l|kpL#=m?M%lQmow_}7+y4Q#F*X@$ z)2m`sD#t>dd*ZWrc&R&WXsTn*X<0-FPBb4 z(@&Q81(r#_iPC5xe^Vpw$oP}gT-jx-vSDP&@-BX;Cv%)~)y1-!UO;>@*$eD;Jh6>Z zna=d07LXi9?S;JpFg9z(#512}R~@*fm+VlSqqd*AOF5I}EbZbCQ=8SX`@74@jQxY( zcW@Z?akB*+P3mX%(dW-MTh%;GS^a70tApgJ-xy?;Q1w6qZS{`?qJQ1PSY2>TU#YI+E9yRD z(b1~5(0Ny?VF7R-RzG=(J}{1)y|dcpE<>!A%3ner7<(*Rrw=l1Yd&hwt98W?MvOD3 zkQy3CZv1`&)eS{y&feLewCqo)scEo0fr6eMxA*K1`F zoW&h(g-!UOUwCTZF$LzIt~kFlMpVG|<>t&Gg!50i>at}$E;$_v@4w3qUBpgl*AK_# zM19m+uT3PBxYx#|UQ*_Z^6t`e9ndA}HyOSka|FpebFrFtgl z(HaW@I~Qe4PP4box`{KlT-CdCp2~TE5sTOetw=_MbiVB~LYW#YCE&ip8?&{Nt!qYy zT_?U~B#q^0#`$LEUuM>zXGSt(-gN0Tt+#iMP~xMJS?R1a>xj|@@gB3EqvAe!frklSFjM?Ws zIj3m3v&0>$G)3$5VDHZKFDb3WBc?}79{twiBlD>Nf5MCvcV+DM4G{~*5w5afhv}0c z+gP$5GI7VAn@7wx!q&H^>WL>5QxcYxYGiDtt|FUlRJbDS+3LKpMATVy=! zi8N-f>kY-VV?9qIroUITL{GyvX%6=*L*>S^&i$JC%X54683Zh@%0igDDsm|l8hx`Qo)1@SNUCKKc6NtNn3&#Inr{a6@?SZqg zLyIaNK?bR%hI^-ST5?n#-AaoEe)^cAcs4_$hL9>k@F>@p&lU;k;m^o|$r7W0? zUxXaNoF~^#l#`K28e()y=#lor*30z&Og<6Wv zcLExc!WX?KM1ZKm3&pxZLaM=5*+IqrZAqQJpzAt4)ZWA1BDx@ZYCyR_QB$_X zcgN)oDzcpf760D?*^=?;v3OAc8}K zc-3+R!>a z-0wU&-g^bU#O`I{bF4iGu$PxFp?mZ9Ikb-t{pW0=?1XuT0Dus0tnpTi2!Hn%>b{g1 literal 0 HcmV?d00001 diff --git a/requirements.txt b/requirements.txt index 3b7c68ca..bdd871f6 100644 --- a/requirements.txt +++ b/requirements.txt @@ -21,3 +21,4 @@ xgboost==2.1.3 numpy==1.26.4 requests==2.32.3 lightgbm==4.5.0 +tabpfn==2.0.3; python_version <= '3.11' diff --git a/shapiq/__init__.py b/shapiq/__init__.py index e3195949..9d483a92 100644 --- a/shapiq/__init__.py +++ b/shapiq/__init__.py @@ -2,7 +2,7 @@ the well established Shapley value and its generalization to interaction. """ -__version__ = "1.1.2" +__version__ = "1.2.0" # approximator classes from .approximator import ( @@ -39,14 +39,14 @@ from .datasets import load_adult_census, load_bike_sharing, load_california_housing # explainer classes -from .explainer import Explainer, TabularExplainer, TreeExplainer +from .explainer import Explainer, TabPFNExplainer, TabularExplainer, TreeExplainer # exact computer classes from .game_theory.exact import ExactComputer # game classes # imputer classes -from .games import BaselineImputer, ConditionalImputer, Game, MarginalImputer +from .games import BaselineImputer, ConditionalImputer, Game, MarginalImputer, TabPFNImputer # base classes from .interaction_values import InteractionValues @@ -97,10 +97,12 @@ "Explainer", "TabularExplainer", "TreeExplainer", + "TabPFNExplainer", # imputers "MarginalImputer", "BaselineImputer", "ConditionalImputer", + "TabPFNImputer", # plots "network_plot", "stacked_bar_plot", diff --git a/shapiq/explainer/__init__.py b/shapiq/explainer/__init__.py index c04f1121..eb8403af 100644 --- a/shapiq/explainer/__init__.py +++ b/shapiq/explainer/__init__.py @@ -1,7 +1,8 @@ """Explainer objects, including TreeSHAP-IQ.""" from ._base import Explainer +from .tabpfn import TabPFNExplainer from .tabular import TabularExplainer from .tree import TreeExplainer -__all__ = ["Explainer", "TabularExplainer", "TreeExplainer"] +__all__ = ["Explainer", "TabularExplainer", "TreeExplainer", "TabPFNExplainer"] diff --git a/shapiq/explainer/_base.py b/shapiq/explainer/_base.py index 3bc9442c..60393813 100644 --- a/shapiq/explainer/_base.py +++ b/shapiq/explainer/_base.py @@ -1,6 +1,8 @@ """The base Explainer classes for the shapiq package.""" +from abc import abstractmethod from typing import Optional +from warnings import warn import numpy as np @@ -12,7 +14,10 @@ class Explainer: """The main Explainer class for a simpler user interface. shapiq.Explainer is a simplified interface for the ``shapiq`` package. It detects between - TabularExplainer and TreeExplainer based on the model class. + :class:`~shapiq.explainer.tabular.TabularExplainer`, + :class:`~shapiq.explainer.tree.TreeExplainer`, + and :class:`~shapiq.explainer.tabpfn.TabPFNExplainer`. For a detailed description of the + different explainers, see the respective classes. Args: model: The model object to be explained. @@ -32,24 +37,14 @@ def __init__( ) -> None: self._model_class = print_class(model) - self._predict_function, self._model_type = get_predict_function_and_model_type( + self._shapiq_predict_function, self._model_type = get_predict_function_and_model_type( model, self._model_class, class_index ) self.model = model if data is not None: - if not isinstance(data, np.ndarray): - raise TypeError("`data` must be a NumPy array.") - try: - pred = self.predict(data) - if isinstance(pred, np.ndarray): - if len(pred.shape) > 1: - raise ValueError() - else: - raise ValueError() - except Exception as e: - print(f"Error: The `data` provided is not compatible with the model. {e}") - pass + if self._model_type != "tabpfn": + self._validate_data(data) self.data = data # not super() @@ -59,13 +54,66 @@ def __init__( self.__class__ = _explainer _explainer.__init__(self, model=model, data=data, class_index=class_index, **kwargs) - def explain(self, x: np.ndarray) -> InteractionValues: - """Explain the model's prediction in terms of interaction values. + def _validate_data(self, data: np.ndarray, raise_error: bool = False) -> None: + """Validate the data for compatibility with the model. Args: - x: An instance/point/sample/observation to be explained. + data: A 2-dimensional matrix of inputs to be explained. + raise_error: Whether to raise an error if the data is not compatible with the model or + only print a warning. Defaults to ``False``. + + Raises: + TypeError: If the data is not a NumPy array. + """ + message = "The `data` and the model must be compatible." + if not isinstance(data, np.ndarray): + message += " The `data` must be a NumPy array." + raise TypeError(message) + try: + # TODO (mmschlk): This can take a long time for large datasets and slow models + pred = self.predict(data) + if isinstance(pred, np.ndarray): + if len(pred.shape) > 1: + message += " The model's prediction must be a 1-dimensional array." + raise ValueError() + else: + message += " The model's prediction must be a NumPy array." + raise ValueError() + except Exception as e: + if raise_error: + raise ValueError(message) from e + else: + warn(message) + + def explain(self, x: np.ndarray, *args, **kwargs) -> InteractionValues: + """Explain a single prediction in terms of interaction values. + + Args: + x: A numpy array of a data point to be explained. + *args: Additional positional arguments passed to the explainer. + **kwargs: Additional keyword-only arguments passed to the explainer. + + Returns: + The interaction values of the prediction. + """ + explanation = self.explain_function(x=x, *args, **kwargs) + if explanation.min_order == 0: + explanation[()] = explanation.baseline_value + return explanation + + @abstractmethod + def explain_function(self, x: np.ndarray, *args, **kwargs) -> InteractionValues: + """Explain a single prediction in terms of interaction values. + + Args: + x: A numpy array of a data point to be explained. + *args: Additional positional arguments passed to the explainer. + **kwargs: Additional keyword-only arguments passed to the explainer. + + Returns: + The interaction values of the prediction. """ - return {} + raise NotImplementedError("The method `explain` must be implemented in a subclass.") def explain_X( self, X: np.ndarray, n_jobs=None, random_state=None, **kwargs @@ -104,4 +152,4 @@ def predict(self, x: np.ndarray) -> np.ndarray: Args: x: An instance/point/sample/observation to be explained. """ - return self._predict_function(self.model, x) + return self._shapiq_predict_function(self.model, x) diff --git a/shapiq/explainer/tabpfn.py b/shapiq/explainer/tabpfn.py new file mode 100644 index 00000000..accffabf --- /dev/null +++ b/shapiq/explainer/tabpfn.py @@ -0,0 +1,120 @@ +"""This module contains the TabPFNExplainer class, which is a class for explaining the predictions +of a TabPFN model.""" + +from typing import Optional, Union + +import numpy as np + +from ..approximator._base import Approximator +from .tabular import TabularExplainer +from .utils import ModelType, get_predict_function_and_model_type + + +class TabPFNExplainer(TabularExplainer): + """The TabPFN explainer as the main interface for the shapiq package. + + The ``TabPFNExplainer`` class is the dedicated interface for the ``shapiq`` package and + TabPFN[2]_ models such as the ``TabPFNClassifier`` and ``TabPFNRegressor``. The explainer + does not rely on classical imputation methods and is optimized for TabPFN's in-context learning + approach. The explanation paradigm for TabPFN is described in Runel et al. (2024)[1]_. In + essence the explainer is a wrapper around the ``TabularExplainer`` class and uses the same API. + + Args: + model: Either a TabPFNClassifier or TabPFNRegressor model to be explained. + + data: The background data to use for the explainer as a 2-dimensional array with shape + ``(n_samples, n_features)``. This data is used to contextualize the model on. + + labels: The labels for the background data as a 1-dimensional array with shape + ``(n_samples,)``. This data is used to contextualize the model on. + + index: The index to explain the model with. Defaults to ``"k-SII"`` which computes the + k-Shapley Interaction Index. If ``max_order`` is set to 1, this corresponds to the + Shapley value (``index="SV"``). Options are: + - ``"SV"``: Shapley value + - ``"k-SII"``: k-Shapley Interaction Index + - ``"FSII"``: Faithful Shapley Interaction Index + - ``"STII"``: Shapley Taylor Interaction Index + - ``"SII"``: Shapley Interaction Index (not recommended for XAI since the values do + not sum up to the prediction) + + x_test: An optional test data set to compute the model's empty prediction (average + prediction) on. If no test data and ``empty_prediction`` is set to ``None`` the last + 20% of the background data is used as test data and the remaining 80% as training data + for contextualization. Defaults to ``None``. + + empty_prediction: Optional value for the model's average prediction on an empty data point + (all features missing). If provided, overrides parameters in ``x_test``. and skips the + computation of the empty prediction. Defaults to ``None``. + + class_index: The class index of the model to explain. Defaults to ``None``, which will set + the class index to ``1`` per default for classification models and is ignored for + regression models. + + approximator: The approximator to use for calculating the Shapley values or Shapley + interactions. Can be a string or an instance of an approximator. Defaults to ``"auto"``. + + verbose: Whether to show a progress bar during the computation. Defaults to ``False``. + Note that verbosity can slow down the computation for large datasets. + + + References: + .. [1] Rundel, D., Kobialka, J., von Crailsheim, C., Feurer, M., Nagler, T., Rügamer, D. (2024). Interpretable Machine Learning for TabPFN. In: Longo, L., Lapuschkin, S., Seifert, C. (eds) Explainable Artificial Intelligence. xAI 2024. Communications in Computer and Information Science, vol 2154. Springer, Cham. https://doi.org/10.1007/978-3-031-63797-1_23 + .. [2] Hollmann, N., Müller, S., Purucker, L. et al. Accurate predictions on small data with a tabular foundation model. Nature 637, 319–326 (2025). https://doi.org/10.1038/s41586-024-08328-6 + """ + + def __init__( + self, + *, + model: ModelType, + data: np.ndarray, + labels: np.ndarray, + index: str = "k-SII", + max_order: int = 2, + x_test: Optional[np.ndarray] = None, + empty_prediction: Optional[float] = None, + class_index: Optional[int] = None, + approximator: Union[str, Approximator] = "auto", + verbose: bool = False, + ): + from ..games.imputer.tabpfn_imputer import TabPFNImputer + + _predict_function, _ = get_predict_function_and_model_type(model, class_index=class_index) + model._shapiq_predict_function = _predict_function + + # check that data and labels have the same number of samples + if data.shape[0] != labels.shape[0]: + raise ValueError( + f"The number of samples in `data` and `labels` must be equal (got data.shape= " + f"{data.shape} and labels.shape={labels.shape})." + ) + n_samples = data.shape[0] + x_train = data + y_train = labels + + if x_test is None and empty_prediction is None: + sections = [int(0.8 * n_samples)] + x_train, x_test = np.split(data, sections) + y_train, _ = np.split(labels, sections) + + if x_test is None: + x_test = x_train # is not used in the TabPFNImputer if empty_prediction is set + + imputer = TabPFNImputer( + model=model, + x_train=x_train, + y_train=y_train, + x_test=x_test, + empty_prediction=empty_prediction, + verbose=verbose, + ) + + super().__init__( + model, + data=x_test, + imputer=imputer, + class_index=class_index, + approximator=approximator, + index=index, + max_order=max_order, + ) diff --git a/shapiq/explainer/tabular.py b/shapiq/explainer/tabular.py index b958ad82..1bfb714a 100644 --- a/shapiq/explainer/tabular.py +++ b/shapiq/explainer/tabular.py @@ -2,6 +2,7 @@ import warnings from typing import Optional, Union +from warnings import warn import numpy as np @@ -49,27 +50,47 @@ class TabularExplainer(Explainer): """The tabular explainer as the main interface for the shapiq package. - The ``TabularExplainer`` class is the main interface for the ``shapiq`` package. It can be used - to explain the predictions of a model by estimating the Shapley interaction values. + The ``TabularExplainer`` class is the main interface for the ``shapiq`` package and tabular + data. It can be used to explain the predictions of any model by estimating the Shapley + interaction values. Args: model: The model to be explained as a callable function expecting data points as input and returning 1-dimensional predictions. + data: A background dataset to be used for imputation. + class_index: The class index of the model to explain. Defaults to ``None``, which will set the class index to ``1`` per default for classification models and is ignored for regression models. + imputer: Either an object of class Imputer or a string from ``["marginal", "conditional"]``. Defaults to ``"marginal"``, which innitializes the default MarginalImputer. - approximator: An approximator object to use for the explainer. Defaults to ``"auto"``, which will - automatically choose the approximator based on the number of features and the number of - samples in the background data. - index: Type of Shapley interaction index to use. Must be one of ``"SII"`` (Shapley Interaction Index), - ``"k-SII"`` (k-Shapley Interaction Index), ``"STII"`` (Shapley-Taylor Interaction Index), - ``"FSII"`` (Faithful Shapley Interaction Index), or ``"SV"`` (Shapley Value) for ``max_order=1``. - Defaults to ``"k-SII"``. - max_order: The maximum interaction order to be computed. Defaults to ``2``. - random_state: The random state to initialize Imputer and Approximator with. Defaults to ``None``. + + approximator: An approximator object to use for the explainer. Defaults to ``"auto"`` + which will automatically choose the approximator based on the number of features and + the desired index. + - for index ``"SV"``: :class:`~shapiq.approximator.KernelSHAP` + - for index ``"SII"`` or ``"k-SII"``: :class:`~shapiq.approximator.KernelSHAPIQ` + - for index ``"FSII"``: :class:`~shapiq.approximator.RegressionFSII` + - for index ``"STII"``: :class:`~shapiq.approximator.SVARMIQ` + + index: The index to explain the model with. Defaults to ``"k-SII"`` which computes the + k-Shapley Interaction Index. If ``max_order`` is set to 1, this corresponds to the + Shapley value (``index="SV"``). Options are: + - ``"SV"``: Shapley value + - ``"k-SII"``: k-Shapley Interaction Index + - ``"FSII"``: Faithful Shapley Interaction Index + - ``"STII"``: Shapley Taylor Interaction Index + - ``"SII"``: Shapley Interaction Index (not recommended for XAI since the values do + not sum up to the prediction) + + max_order: The maximum interaction order to be computed. Defaults to ``2``. Set to ``1`` for + no interactions (single feature importance). + + random_state: The random state to initialize Imputer and Approximator with. Defaults to + ``None``. + **kwargs: Additional keyword-only arguments passed to the imputer. Attributes: @@ -90,15 +111,31 @@ def __init__( index: str = "k-SII", max_order: int = 2, random_state: Optional[int] = None, + verbose: bool = False, **kwargs, ) -> None: - from shapiq.games.imputer import BaselineImputer, ConditionalImputer, MarginalImputer + from shapiq.games.imputer import ( + BaselineImputer, + ConditionalImputer, + MarginalImputer, + TabPFNImputer, + ) if index not in AVAILABLE_INDICES: raise ValueError(f"Invalid index `{index}`. " f"Valid indices are {AVAILABLE_INDICES}.") super().__init__(model, data, class_index) + # get class for self + class_name = self.__class__.__name__ + if self._model_type == "tabpfn" and class_name == "TabularExplainer": + warn( + "You are using a TabPFN model with the ``shapiq.TabularExplainer`` directly. This " + "is not recommended as it uses missing value imputation and not contextualization. " + "Consider using the ``shapiq.TabPFNExplainer`` instead. For more information see " + "the documentation and the example notebooks." + ) + self._random_state = random_state if imputer == "marginal": self._imputer = MarginalImputer( @@ -116,20 +153,23 @@ def __init__( isinstance(imputer, MarginalImputer) or isinstance(imputer, ConditionalImputer) or isinstance(imputer, BaselineImputer) + or isinstance(imputer, TabPFNImputer) ): self._imputer = imputer else: raise ValueError( f"Invalid imputer {imputer}. " - f'Must be one of ["marginal", "conditional"], or a valid Imputer object.' + f'Must be one of ["marginal", "baseline", "conditional"], or a valid Imputer ' + f"object." ) self._n_features: int = self.data.shape[1] + self._imputer.verbose = verbose # set the verbose flag for the imputer self.index = index self._max_order: int = max_order self._approximator = self._init_approximator(approximator, self.index, self._max_order) - def explain( + def explain_function( self, x: np.ndarray, budget: Optional[int] = None, random_state: Optional[int] = None ) -> InteractionValues: """Explains the model's predictions. @@ -178,7 +218,8 @@ def _init_approximator( if max_order == 1: if index != "SV": warnings.warn( - "`max_order=1` but `index != 'SV'`, setting `index = 'SV'`. Using the KernelSHAP approximator." + "`max_order=1` but `index != 'SV'`, setting `index = 'SV'`. " + "Using the KernelSHAP approximator." ) self.index = "SV" return KernelSHAP( @@ -188,7 +229,8 @@ def _init_approximator( elif index == "SV": if max_order != 1: warnings.warn( - "`index='SV'` but `max_order != 1`, setting `max_order = 1`. Using the KernelSHAP approximator." + "`index='SV'` but `max_order != 1`, setting `max_order = 1`. " + "Using the KernelSHAP approximator." ) self._max_order = 1 return KernelSHAP( diff --git a/shapiq/explainer/tree/explainer.py b/shapiq/explainer/tree/explainer.py index 7e172f26..8dff6fbe 100644 --- a/shapiq/explainer/tree/explainer.py +++ b/shapiq/explainer/tree/explainer.py @@ -75,7 +75,16 @@ def __init__( ] self.baseline_value = self._compute_baseline_value() - def explain(self, x: np.ndarray) -> InteractionValues: + def explain_function(self, x: np.ndarray, **kwargs) -> InteractionValues: + """Computes the Shapley Interaction values for a single instance. + + Args: + x: The instance to explain as a 1-dimensional array. + **kwargs: Additional keyword arguments are ignored. + + Returns: + The interaction values for the instance. + """ if len(x.shape) != 1: raise TypeError("explain expects a single instance, not a batch.") # run treeshapiq for all trees diff --git a/shapiq/explainer/utils.py b/shapiq/explainer/utils.py index 576a4518..42ed2c4a 100644 --- a/shapiq/explainer/utils.py +++ b/shapiq/explainer/utils.py @@ -21,17 +21,42 @@ def get_explainers() -> dict[str, Any]: Returns: A dictionary of all available explainer classes. """ + from shapiq.explainer.tabpfn import TabPFNExplainer from shapiq.explainer.tabular import TabularExplainer from shapiq.explainer.tree.explainer import TreeExplainer - return {"tabular": TabularExplainer, "tree": TreeExplainer} + return {"tabular": TabularExplainer, "tree": TreeExplainer, "tabpfn": TabPFNExplainer} def get_predict_function_and_model_type( - model: ModelType, model_class: str, class_index: Optional[int] = None + model: ModelType, + model_class: Optional[str] = None, + class_index: Optional[int] = None, ) -> tuple[Callable[[ModelType, np.ndarray], np.ndarray], str]: + """Get the predict function and model type for a given model. + + The prediction function is used in the explainer to predict the model's output for a given data + point. The function has the following signature: ``predict_function(model, data)``. + + Args: + model: The model to explain. Can be any model object or callable function. We try to infer + the model type from the model object. + + model_class: The class of the model. as a string. If not provided, it will be inferred from + the model object. + + class_index: The class index of the model to explain. Defaults to ``None``, which will set + the class index to ``1`` per default for classification models and is ignored for + regression models. + + Returns: + A tuple of the predict function and the model type. + """ from . import tree + if model_class is None: + model_class = print_class(model) + _model_type = "tabular" # default _predict_function = None @@ -96,6 +121,12 @@ def get_predict_function_and_model_type( _model_type = "tabular" _predict_function = predict_tensorflow + if model_class in [ + "tabpfn.classifier.TabPFNClassifier", + "tabpfn.regressor.TabPFNRegressor", + ]: + _model_type = "tabpfn" + # default extraction (sklearn api) if _predict_function is None and hasattr(model, "predict_proba"): _predict_function = predict_proba diff --git a/shapiq/games/__init__.py b/shapiq/games/__init__.py index bb5f4ddb..876d3077 100644 --- a/shapiq/games/__init__.py +++ b/shapiq/games/__init__.py @@ -2,8 +2,8 @@ # from . import benchmark # not imported here to avoid circular imports and long import times from .base import Game -from .imputer import BaselineImputer, ConditionalImputer, MarginalImputer +from .imputer import BaselineImputer, ConditionalImputer, MarginalImputer, TabPFNImputer -__all__ = ["Game", "MarginalImputer", "ConditionalImputer", "BaselineImputer"] +__all__ = ["Game", "MarginalImputer", "ConditionalImputer", "BaselineImputer", "TabPFNImputer"] # Path: shapiq/games/__init__.py diff --git a/shapiq/games/imputer/__init__.py b/shapiq/games/imputer/__init__.py index 952a741d..e48d4b23 100644 --- a/shapiq/games/imputer/__init__.py +++ b/shapiq/games/imputer/__init__.py @@ -3,5 +3,6 @@ from .baseline_imputer import BaselineImputer from .conditional_imputer import ConditionalImputer from .marginal_imputer import MarginalImputer +from .tabpfn_imputer import TabPFNImputer -__all__ = ["MarginalImputer", "ConditionalImputer", "BaselineImputer"] +__all__ = ["MarginalImputer", "ConditionalImputer", "BaselineImputer", "TabPFNImputer"] diff --git a/shapiq/games/imputer/base.py b/shapiq/games/imputer/base.py index 52c2b88b..b58b73ca 100644 --- a/shapiq/games/imputer/base.py +++ b/shapiq/games/imputer/base.py @@ -15,15 +15,23 @@ class Imputer(Game): Args: model: The model to explain as a callable function expecting a data points as input and returning the model's predictions. + data: The background data to use for the explainer as a 2-dimensional array with shape ``(n_samples, n_features)``. + x: The explanation point to use the imputer on either as a 2-dimensional array with shape ``(1, n_features)`` or as a vector with shape ``(n_features,)``. + sample_size: The number of samples to draw from the background data. Defaults to ``100`` but can is usually overwritten in the subclasses. + categorical_features: A list of indices of the categorical features in the background data. + random_state: The random state to use for sampling. Defaults to ``None``. + verbose: A flag to enable verbose imputation, which will print a progress bar for model + evaluation. Note that this can slow down the imputation process. Defaults to ``False``. + Attributes: n_features: The number of features in the data (equals the number of players in the game). data: The background data to use for the imputer. @@ -45,11 +53,15 @@ def __init__( sample_size: int = 100, categorical_features: list[int] = None, random_state: Optional[int] = None, + verbose: bool = False, ) -> None: if callable(model) and not hasattr(model, "_predict_function"): self._predict_function = utils.predict_callable - else: # shapiq.Explainer adds a predict function to the model to make it callable - self._predict_function = model._predict_function + # shapiq.Explainer adds a _shapiq_predict_function to the model to make it callable + elif hasattr(model, "_shapiq_predict_function"): + self._predict_function = model._shapiq_predict_function + else: + raise ValueError("The model must be callable or have a predict function.") self.model = model # check if data is a vector if data.ndim == 1: @@ -69,7 +81,7 @@ def __init__( # init the game # developer note: the normalization_value needs to be set in the subclass - super().__init__(n_players=self.n_features, normalize=False) + super().__init__(n_players=self.n_features, normalize=False, verbose=verbose) @property def x(self) -> Optional[np.ndarray]: diff --git a/shapiq/games/imputer/tabpfn_imputer.py b/shapiq/games/imputer/tabpfn_imputer.py new file mode 100644 index 00000000..b3fcf768 --- /dev/null +++ b/shapiq/games/imputer/tabpfn_imputer.py @@ -0,0 +1,110 @@ +"""This module contains the TabPFNImputer class, which incorporates the Remove-and-Contextualize +paradigm of explaining the TabPFN model's predictions.""" + +from typing import Callable, Optional + +import numpy as np + +from ...explainer.utils import ModelType +from .base import Imputer + + +class TabPFNImputer(Imputer): + """An Imputer for TabPFN using the Remove-and-Contextualize paradigm. + + The remove-and-contextualize paradigm is a strategy to explain the predictions of a TabPFN[2]_ + model which uses in-context learning for prediction. Instead of imputing missing features, the + TabPFNImputer removes feature columns missing in a coalition from training data and re-"trains" + re-contextualizes the model with the remaining features. The model is then used to predict the + data point which is also missing the features. This pardigm is described in Rundel et al. + (2024)[1]_. + + Args: + model: The model to be explained as a callable function expecting data points as input and + returning 1-dimensional predictions. + + x_train: The training data to "train" the model on. Note that the model is not actually + trained but the correct train data with the correct features per coalition are put into + TabPFN's context. + + y_train: The training labels to "train" the model on. Note that the model is not actually + trained but the correct train data and labels are put into TabPFN's context. + + x_test: The test data to evaluate the model's average (empty) prediction on. If no test + data is provided, the empty prediction must be given. Defaults to ``None``. + + empty_prediction: The model's average prediction on an empty data point (all features + missing). This can be computed by averaging the model's predictions on the test data. + + Attributes: + x_train: The training data to contextualize the model on. + y_train: The training labels to contextualize the model on. + empty_prediction: The model's average prediction on an empty data point. + + References: + .. [1] Rundel, D., Kobialka, J., von Crailsheim, C., Feurer, M., Nagler, T., Rügamer, D. (2024). Interpretable Machine Learning for TabPFN. In: Longo, L., Lapuschkin, S., Seifert, C. (eds) Explainable Artificial Intelligence. xAI 2024. Communications in Computer and Information Science, vol 2154. Springer, Cham. https://doi.org/10.1007/978-3-031-63797-1_23 + .. [2] Hollmann, N., Müller, S., Purucker, L. et al. Accurate predictions on small data with a tabular foundation model. Nature 637, 319–326 (2025). https://doi.org/10.1038/s41586-024-08328-6 + + """ + + def __init__( + self, + model: ModelType, + x_train: np.ndarray, + y_train: np.ndarray, + x_test: Optional[np.ndarray] = None, + empty_prediction: Optional[float] = None, + verbose: bool = False, + predict_function: Optional[Callable[[ModelType, np.ndarray], np.ndarray]] = None, + ): + self.x_train = x_train + self.y_train = y_train + + if not hasattr(model, "_shapiq_predict_function"): + if predict_function is None: + raise ValueError( + f"If the Imputer is not instantiated via a ``shapiq.Explainer`` object, you" + f" must provide a ``predict_function`` (received" + f" predict_function={predict_function})." + ) + model._shapiq_predict_function = predict_function + + if x_test is None and empty_prediction is None: + raise ValueError("The empty prediction must be given if no test data is provided") + + super().__init__( + model=model, data=x_test, x=None, sample_size=None, random_state=None, verbose=verbose + ) + + if empty_prediction is None: + self.model.fit(x_train, y_train) # contextualize the model on the training data + predictions = self.predict(x_test) + empty_prediction = np.mean(predictions) + self.empty_prediction = empty_prediction + + def value_function(self, coalitions: np.ndarray) -> np.ndarray: + """The value function performs the remove-and-contextualize strategy for TabPFN. + + The value function removes absent features from a coalition by "training" the model again + on the subset of features. The model is then used to predict the data point with the + missing features. + + Args: + coalitions: A boolean array indicating which features are present (``True``) and which + are missing (``False``). The shape of the array must be ``(n_subsets, n_players)``. + + Returns: + The model's predictions on the restricted data points. The shape of the array is + ``(n_subsets,)``. + """ + output = np.zeros(len(coalitions), dtype=float) + for i, coalition in enumerate(coalitions): + if sum(coalition) == 0: + output[i] = self.empty_prediction + continue + x_train_coal = self.x_train[:, coalition] + x_explain_coal = self.x[:, coalition] + self.model.fit(x_train_coal, self.y_train) + pred = float(self.predict(x_explain_coal)) + output[i] = pred + return output diff --git a/shapiq/interaction_values.py b/shapiq/interaction_values.py index cb1fb931..0b619973 100644 --- a/shapiq/interaction_values.py +++ b/shapiq/interaction_values.py @@ -80,7 +80,19 @@ def __post_init__(self) -> None: ) if not isinstance(self.baseline_value, (int, float)): - raise TypeError("Baseline value must be provided as a number.") + raise TypeError( + f"Baseline value must be provided as a number. Got {self.baseline_value}." + ) + + # check if () is in the interaction_lookup if min_order is 0 -> add it to the end + if self.min_order == 0 and () not in self.interaction_lookup: + self.interaction_lookup[()] = len(self.interaction_lookup) + self.values = np.concatenate((self.values, np.array([self.baseline_value]))) + + # update the baseline value in the values vector if index is not SII + # # TODO: this might be a good idea check if this is okay to do + # if self.index != "SII" and self.baseline_value != self.values[self.interaction_lookup[()]]: + # self.values[self.interaction_lookup[()]] = self.baseline_value @property def dict_values(self) -> dict[tuple[int, ...], float]: @@ -226,6 +238,28 @@ def __getitem__(self, item: Union[int, tuple[int, ...]]) -> float: except KeyError: return 0.0 + def __setitem__(self, item: Union[int, tuple[int, ...]], value: float) -> None: + """Sets the score for the given interaction. + + Args: + item: The interaction as a tuple of integers for which to set the score. If ``item`` is an + integer it serves as the index to the values vector. + value: The value to set for the interaction. + + Raises: + KeyError: If the interaction is not found in the InteractionValues object. + """ + try: + if isinstance(item, int): + self.values[item] = value + else: + item = tuple(sorted(item)) + self.values[self.interaction_lookup[item]] = value + except Exception as e: + raise KeyError( + f"Interaction {item} not found in the InteractionValues. Unable to set a value." + ) from e + def __eq__(self, other: object) -> bool: """Checks if two InteractionValues objects are equal. diff --git a/tests/conftest.py b/tests/conftest.py index b8c5a40e..585bc81b 100644 --- a/tests/conftest.py +++ b/tests/conftest.py @@ -4,6 +4,7 @@ """ import os +from typing import Any import numpy as np import pytest @@ -11,6 +12,7 @@ from sklearn.datasets import make_classification, make_regression from sklearn.ensemble import IsolationForest, RandomForestClassifier, RandomForestRegressor from sklearn.linear_model import LinearRegression, LogisticRegression +from sklearn.model_selection import train_test_split from sklearn.tree import DecisionTreeClassifier, DecisionTreeRegressor NR_FEATURES = 7 @@ -50,6 +52,30 @@ def value_function(self, coalitions: np.ndarray) -> np.ndarray: return CookingGame() +@pytest.fixture +def tabpfn_classification_problem() -> tuple[Any, np.ndarray, np.ndarray, np.ndarray]: + """Returns a very simple tabpfn classifier and dataset.""" + from tabpfn import TabPFNClassifier + + data, labels = make_classification(n_samples=10, n_features=3, random_state=42, n_redundant=1) + data, x_test, labels, _ = train_test_split(data, labels, random_state=42, train_size=8) + model = TabPFNClassifier() + model.fit(data, labels) + return model, data, labels, x_test + + +@pytest.fixture +def tabpfn_regression_problem() -> tuple[Any, np.ndarray, np.ndarray, np.ndarray]: + """Returns a very simple tabpfn regressor and dataset.""" + from tabpfn import TabPFNRegressor + + data, labels = make_regression(n_samples=10, n_features=3, random_state=42) + data, x_test, labels, _ = train_test_split(data, labels, random_state=42, train_size=8) + model = TabPFNRegressor() + model.fit(data, labels) + return model, data, labels, x_test + + @pytest.fixture def dt_reg_model() -> DecisionTreeRegressor: """Return a simple decision tree model.""" diff --git a/tests/requirements/requirements.txt b/tests/requirements/requirements.txt index 7a750393..5bc8471a 100644 --- a/tests/requirements/requirements.txt +++ b/tests/requirements/requirements.txt @@ -21,3 +21,4 @@ requests==2.32.3 lightgbm==4.5.0 tf-keras==2.18.0 tensorflow==2.18.0 +tabpfn==2.0.3; python_version <= '3.11' diff --git a/tests/test_base_interaction_values.py b/tests/test_base_interaction_values.py index a93f066f..373dafeb 100644 --- a/tests/test_base_interaction_values.py +++ b/tests/test_base_interaction_values.py @@ -112,6 +112,18 @@ def test_initialization(index, n, min_order, max_order, estimation_budget, estim assert interaction_values[0] == interaction_values.values[0] assert interaction_values[-1] == interaction_values.values[-1] + # check setitem + interaction_values[(0,)] = 999_999 + assert interaction_values[(0,)] == 999_999 + + # check setitem with integer as input + interaction_values[0] = 111_111 + assert interaction_values[0] == 111_111 + + # check setitem raises error for invalid interaction + with pytest.raises(KeyError): + interaction_values[(100, 101)] = 0 + # test __len__ assert len(interaction_values) == len(interaction_values.values) diff --git a/tests/tests_explainer/test_explainer_models.py b/tests/tests_explainer/test_explainer_models.py index 3d48feed..f718614a 100644 --- a/tests/tests_explainer/test_explainer_models.py +++ b/tests/tests_explainer/test_explainer_models.py @@ -17,7 +17,7 @@ def test_torch_reg(torch_reg_model, background_reg_data): explainer = Explainer(model=torch_reg_model, data=background_reg_data) values = explainer.explain(x_explain) assert isinstance(values, InteractionValues) - sum_of_values = sum(values.values) + values.baseline_value + sum_of_values = sum(values.values) assert prediction == pytest.approx(sum_of_values, rel=0.01) @@ -32,13 +32,13 @@ def test_torch_clf(torch_clf_model, background_clf_data): explainer = Explainer(model=torch_clf_model, data=background_clf_data, class_index=2) values = explainer.explain(x_explain) assert isinstance(values, InteractionValues) - sum_of_values = sum(values.values) + values.baseline_value + sum_of_values = sum(values.values) assert prediction[2] == pytest.approx(sum_of_values, rel=0.001) explainer = Explainer(model=torch_clf_model, data=background_clf_data, class_index=0) values = explainer.explain(x_explain) assert isinstance(values, InteractionValues) - sum_of_values = sum(values.values) + values.baseline_value + sum_of_values = sum(values.values) assert prediction[0] == pytest.approx(sum_of_values, rel=0.001) @@ -51,13 +51,13 @@ def test_sklearn_clf_tree(dt_clf_model, background_clf_data): explainer = TabularExplainer(model=dt_clf_model, data=background_clf_data, class_index=2) values = explainer.explain(x_explain) assert isinstance(values, InteractionValues) - sum_of_values = sum(values.values) + values.baseline_value + sum_of_values = sum(values.values) assert prediction[2] == pytest.approx(sum_of_values, abs=0.001) explainer = TabularExplainer(model=dt_clf_model, data=background_clf_data, class_index=0) values = explainer.explain(x_explain) assert isinstance(values, InteractionValues) - sum_of_values = sum(values.values) + values.baseline_value + sum_of_values = sum(values.values) assert prediction[0] == pytest.approx(sum_of_values, abs=0.001) # do the same with the bare explainer (only for class_label=2) @@ -78,7 +78,7 @@ def test_sklearn_reg_tree(dt_reg_model, background_reg_data): explainer = TabularExplainer(model=dt_reg_model, data=background_reg_data) values = explainer.explain(x_explain) assert isinstance(values, InteractionValues) - sum_of_values = sum(values.values) + values.baseline_value + sum_of_values = sum(values.values) assert prediction == pytest.approx(sum_of_values, abs=0.001) # do the same with the bare explainer @@ -99,13 +99,13 @@ def test_sklearn_clf_forest(rf_clf_model, background_clf_data): explainer = TabularExplainer(model=rf_clf_model, data=background_clf_data, class_index=2) values = explainer.explain(x_explain) assert isinstance(values, InteractionValues) - sum_of_values = sum(values.values) + values.baseline_value + sum_of_values = sum(values.values) assert prediction[2] == pytest.approx(sum_of_values, rel=0.001) explainer = TabularExplainer(model=rf_clf_model, data=background_clf_data, class_index=0) values = explainer.explain(x_explain) assert isinstance(values, InteractionValues) - sum_of_values = sum(values.values) + values.baseline_value + sum_of_values = sum(values.values) assert prediction[0] == pytest.approx(sum_of_values, rel=0.001) # do the same with the bare explainer (only for class_label=2) @@ -125,14 +125,14 @@ def test_sklearn_reg_forest(rf_reg_model, background_reg_data): explainer = TabularExplainer(model=rf_reg_model, data=background_reg_data) values = explainer.explain(x_explain) assert isinstance(values, InteractionValues) - sum_of_values = sum(values.values) + values.baseline_value + sum_of_values = sum(values.values) assert prediction == pytest.approx(sum_of_values) # do the same with the bare explainer explainer = Explainer(model=rf_reg_model, data=background_reg_data) values = explainer.explain(x_explain) assert isinstance(values, InteractionValues) - sum_of_values = sum(values.values) + values.baseline_value + sum_of_values = sum(values.values) assert prediction == pytest.approx(sum_of_values, rel=0.01) @@ -145,20 +145,20 @@ def test_sklearn_clf_logistic_regression(lr_clf_model, background_clf_data): explainer = TabularExplainer(model=lr_clf_model, data=background_clf_data, class_index=2) values = explainer.explain(x_explain) assert isinstance(values, InteractionValues) - sum_of_values = sum(values.values) + values.baseline_value + sum_of_values = sum(values.values) assert prediction[2] == pytest.approx(sum_of_values) explainer = TabularExplainer(model=lr_clf_model, data=background_clf_data, class_index=0) values = explainer.explain(x_explain) assert isinstance(values, InteractionValues) - sum_of_values = sum(values.values) + values.baseline_value + sum_of_values = sum(values.values) assert prediction[0] == pytest.approx(sum_of_values) # do the same with the bare explainer (only for class_label=2) explainer = Explainer(model=lr_clf_model, data=background_clf_data, class_index=2) values = explainer.explain(x_explain) assert isinstance(values, InteractionValues) - sum_of_values = sum(values.values) + values.baseline_value + sum_of_values = sum(values.values) assert prediction[2] == pytest.approx(sum_of_values) @@ -171,14 +171,14 @@ def test_sklearn_reg_linear_regression(lr_reg_model, background_reg_data): explainer = TabularExplainer(model=lr_reg_model, data=background_reg_data) values = explainer.explain(x_explain) assert isinstance(values, InteractionValues) - sum_of_values = sum(values.values) + values.baseline_value + sum_of_values = sum(values.values) assert prediction == pytest.approx(sum_of_values) # do the same with the bare explainer explainer = Explainer(model=lr_reg_model, data=background_reg_data) values = explainer.explain(x_explain) assert isinstance(values, InteractionValues) - sum_of_values = sum(values.values) + values.baseline_value + sum_of_values = sum(values.values) assert prediction == pytest.approx(sum_of_values) @@ -191,7 +191,7 @@ def test_lightgbm_reg(lightgbm_reg_model, background_reg_data): explainer = TabularExplainer(model=lightgbm_reg_model, data=background_reg_data) values = explainer.explain(x_explain) assert isinstance(values, InteractionValues) - sum_of_values = sum(values.values) + values.baseline_value + sum_of_values = sum(values.values) assert prediction == pytest.approx(sum_of_values) # do the same with the bare explainer @@ -212,13 +212,13 @@ def test_lightgbm_clf(lightgbm_clf_model, background_clf_data): explainer = TabularExplainer(model=lightgbm_clf_model, data=background_clf_data, class_index=2) values = explainer.explain(x_explain) assert isinstance(values, InteractionValues) - sum_of_values = sum(values.values) + values.baseline_value + sum_of_values = sum(values.values) assert prediction[2] == pytest.approx(sum_of_values, rel=0.001) explainer = TabularExplainer(model=lightgbm_clf_model, data=background_clf_data, class_index=0) values = explainer.explain(x_explain) assert isinstance(values, InteractionValues) - sum_of_values = sum(values.values) + values.baseline_value + sum_of_values = sum(values.values) assert prediction[0] == pytest.approx(sum_of_values, rel=0.001) # do the same with the bare explainer (only for class_label=2) @@ -241,7 +241,7 @@ def test_isoforest_clf(if_clf_model, if_clf_dataset): explainer = TabularExplainer(model=if_clf_model, data=x_data, class_index=2) values = explainer.explain(x_explain) assert isinstance(values, InteractionValues) - sum_of_values = sum(values.values) + values.baseline_value + sum_of_values = sum(values.values) assert pytest.approx(sum_of_values, abs=0.001) == prediction # do the same with the bare explainer diff --git a/tests/tests_explainer/test_explainer_tabular.py b/tests/tests_explainer/test_explainer_tabular.py index 0d5f33bf..d8837fba 100644 --- a/tests/tests_explainer/test_explainer_tabular.py +++ b/tests/tests_explainer/test_explainer_tabular.py @@ -169,8 +169,8 @@ def test_explain(dt_model, data, index, budget, max_order, imputer): # test for efficiency if index in ("FSII", "k-SII"): prediction = float(model_function(x)[0]) - sum_of_values = float(np.sum(interaction_values.values) + interaction_values.baseline_value) - assert interaction_values[()] == 0.0 + sum_of_values = float(np.sum(interaction_values.values)) + assert pytest.approx(interaction_values[()]) == interaction_values.baseline_value assert pytest.approx(sum_of_values, 0.01) == prediction diff --git a/tests/tests_explainer/test_tabpfn_explainer.py b/tests/tests_explainer/test_tabpfn_explainer.py new file mode 100644 index 00000000..c0f0bc6e --- /dev/null +++ b/tests/tests_explainer/test_tabpfn_explainer.py @@ -0,0 +1,61 @@ +"""This test module tests the TabPFNExplainer object.""" + +import sys + +import pytest + +from shapiq import Explainer, InteractionValues, TabPFNExplainer, TabularExplainer + + +@pytest.mark.skipif(sys.version_info > (3, 11), reason="requires python3.11 or lower") +def test_tabpfn_explainer_clf(tabpfn_classification_problem): + """Test the TabPFNExplainer class for classification problems.""" + import tabpfn + + # setup + model, data, labels, x_test = tabpfn_classification_problem + x_explain = x_test[0] + assert isinstance(model, tabpfn.TabPFNClassifier) + if model.n_features_in_ == data.shape[1]: + model.fit(data, labels) + assert model.n_features_in_ == data.shape[1] + + explainer = TabPFNExplainer(model=model, data=data, labels=labels, x_test=x_test) + explanation = explainer.explain(x=x_explain) + assert isinstance(explanation, InteractionValues) + + # test that bare explainer gets turned into TabPFNExplainer + explainer = Explainer(model=model, data=data, labels=labels, x_test=x_test) + assert isinstance(explainer, TabPFNExplainer) + + # test that TabularExplainer works as well + with pytest.warns(UserWarning): + explainer = TabularExplainer(model=model, data=data, class_index=1, imputer="baseline") + assert isinstance(explainer, TabularExplainer) + + +@pytest.mark.skipif(sys.version_info > (3, 11), reason="requires python3.11 or lower") +def test_tabpfn_explainer_reg(tabpfn_regression_problem): + """Test the TabPFNExplainer class for regression problems.""" + import tabpfn + + # setup + model, data, labels, x_test = tabpfn_regression_problem + x_explain = x_test[0] + assert isinstance(model, tabpfn.TabPFNRegressor) + if model.n_features_in_ == data.shape[1]: + model.fit(data, labels) + assert model.n_features_in_ == data.shape[1] + + explainer = TabPFNExplainer(model=model, data=data, labels=labels, x_test=x_test) + explanation = explainer.explain(x=x_explain) + assert isinstance(explanation, InteractionValues) + + # test that bare explainer gets turned into TabPFNExplainer + explainer = Explainer(model=model, data=data, labels=labels, x_test=x_test) + assert isinstance(explainer, TabPFNExplainer) + + # test that TabularExplainer works as well + with pytest.warns(UserWarning): + explainer = TabularExplainer(model=model, data=data, class_index=1, imputer="baseline") + assert isinstance(explainer, TabularExplainer) diff --git a/tests/tests_imputer/test_tabpfn_imputer.py b/tests/tests_imputer/test_tabpfn_imputer.py new file mode 100644 index 00000000..ea64ef01 --- /dev/null +++ b/tests/tests_imputer/test_tabpfn_imputer.py @@ -0,0 +1,100 @@ +"""This test module tests the tabpfn imputer object.""" + +import sys + +import numpy as np +import pytest + +from shapiq import TabPFNImputer +from shapiq.explainer.utils import get_predict_function_and_model_type + + +@pytest.mark.skipif(sys.version_info > (3, 11), reason="requires python3.11 or lower") +def test_tabpfn_imputer(tabpfn_classification_problem): + """Test the TabPFNImputer class.""" + import tabpfn + + # setup + model, data, labels, x_test = tabpfn_classification_problem + assert isinstance(model, tabpfn.TabPFNClassifier) + if model.n_features_in_ == data.shape[1]: + model.fit(data, labels) + assert model.n_features_in_ == data.shape[1] + assert not hasattr(model, "_shapiq_predict_function") + + # setup the tabpfn imputer + prediction_function, _ = get_predict_function_and_model_type(model) + imputer = TabPFNImputer( + model=model, + x_train=data, + y_train=labels, + x_test=x_test, + predict_function=prediction_function, + ) + imputer.fit(x=x_test[0]) + + # test the imputer + imputer(np.asarray([True, True, True])) # 3 features should now been fitted + assert model.n_features_in_ == 3 + imputer(np.asarray([True, True, False])) # 2 features should now been fitted + assert model.n_features_in_ == 2 + imputer(np.asarray([False, True, False])) # 1 feature should now been fitted + assert model.n_features_in_ == 1 + + +@pytest.mark.skipif(sys.version_info > (3, 11), reason="requires python3.11 or lower") +def test_empty_prediction(tabpfn_classification_problem): + """Tests the TabPFNImputer with a manual empty prediction.""" + import tabpfn + + # setup + model, data, labels, x_test = tabpfn_classification_problem + assert isinstance(model, tabpfn.TabPFNClassifier) + if model.n_features_in_ == data.shape[1]: + model.fit(data, labels) + assert model.n_features_in_ == data.shape[1] + assert not hasattr(model, "_shapiq_predict_function") + + manual_empty_prediction = 1000 + + # setup the tabpfn imputer + prediction_function, _ = get_predict_function_and_model_type(model) + imputer = TabPFNImputer( + model=model, + x_train=data, + y_train=labels, + x_test=x_test, + predict_function=prediction_function, + empty_prediction=manual_empty_prediction, + ) + + output = imputer(np.asarray([False, False, False])) + assert output[0] == manual_empty_prediction + + +@pytest.mark.skipif(sys.version_info > (3, 11), reason="requires python3.11 or lower") +def test_tabpfn_imputer_validation(tabpfn_classification_problem): + """Test that the TabPFNImputer raises a ValueError if no predict function is provided.""" + import tabpfn + + # setup + model, data, labels, x_test = tabpfn_classification_problem + assert isinstance(model, tabpfn.TabPFNClassifier) + if model.n_features_in_ == data.shape[1]: + model.fit(data, labels) + assert model.n_features_in_ == data.shape[1] + assert not hasattr(model, "_shapiq_predict_function") + + # no prediction function + with pytest.raises(ValueError): + _ = TabPFNImputer( + model=model, x_train=data, y_train=labels, x_test=x_test, predict_function=None + ) + + # no x_test and no empty prediction + with pytest.raises(ValueError): + + def pred_fun(model, x): + return model.predict_proba(x)[0] + + _ = TabPFNImputer(model=model, x_train=data, y_train=labels, predict_function=pred_fun)