Skip to content

Commit

Permalink
Merge branch 'master' into github_master
Browse files Browse the repository at this point in the history 
Add a notebook for model testing.
  • Loading branch information
@lishen
lishen committed Sep 1, 2017
2 parents d6b3bc7 + 05b0525 commit 114e2ed
Showing 1 changed file with 306 additions and 0 deletions.
306 changes: 306 additions & 0 deletions ddsm_train/Example_model_test.ipynb
View file
Original file line number Diff line number Diff line change
@@ -0,0 +1,306 @@
{
"cells": [
{
"cell_type": "code",
"execution_count": 1,
"metadata": {
"collapsed": false
},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"Populating the interactive namespace from numpy and matplotlib\n"
]
},
{
"name": "stderr",
"output_type": "stream",
"text": [
"Using TensorFlow backend.\n"
]
}
],
"source": [
"%pylab inline\n",
"import pandas as pd\n",
"import numpy as np\n",
"import keras"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Resnet & VGG16"
]
},
{
"cell_type": "code",
"execution_count": 2,
"metadata": {
"collapsed": true
},
"outputs": [],
"source": [
"from keras.models import load_model\n",
"res_mod = load_model('1152x896_s10_prt_addtopf2_1.h5', compile=False)\n",
"vgg_mod = load_model('1152x896_s10_prt_addvgg20.h5', compile=False)\n",
"hybrid_mod = load_model('1152x896_s10_prt_addvgg53.h5', compile=False)"
]
},
{
"cell_type": "code",
"execution_count": 3,
"metadata": {
"collapsed": false
},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"Found 376 images belonging to 2 classes.\n"
]
}
],
"source": [
"from dm_image import DMImageDataGenerator\n",
"test_imgen = DMImageDataGenerator(featurewise_center=True)\n",
"test_imgen.mean = 52.18\n",
"test_generator = test_imgen.flow_from_directory(\n",
" './full_test_1152x896', target_size=(1152, 896), target_scale=None,\n",
" rescale_factor=0.003891,\n",
" equalize_hist=False, dup_3_channels=True, \n",
" classes=['neg', 'pos'], class_mode='categorical', batch_size=4, \n",
" shuffle=False)"
]
},
{
"cell_type": "code",
"execution_count": 4,
"metadata": {
"collapsed": false
},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"[ 0.86367945 0.86367945]\n"
]
}
],
"source": [
"from dm_keras_ext import DMAucModelCheckpoint\n",
"res_auc, res_y_true, res_y_pred = DMAucModelCheckpoint.calc_test_auc(\n",
" test_generator, res_mod, test_samples=test_generator.nb_sample, return_y_res=True)\n",
"print res_auc"
]
},
{
"cell_type": "code",
"execution_count": 5,
"metadata": {
"collapsed": false
},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"[ 0.82536263 0.82536263]\n"
]
}
],
"source": [
"from dm_keras_ext import DMAucModelCheckpoint\n",
"vgg_auc, vgg_y_true, vgg_y_pred = DMAucModelCheckpoint.calc_test_auc(\n",
" test_generator, vgg_mod, test_samples=test_generator.nb_sample, return_y_res=True)\n",
"print vgg_auc"
]
},
{
"cell_type": "code",
"execution_count": 6,
"metadata": {
"collapsed": false
},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"[ 0.84914548 0.84914548]\n"
]
}
],
"source": [
"from dm_keras_ext import DMAucModelCheckpoint\n",
"hybrid_auc, hybrid_y_true, hybrid_y_pred = DMAucModelCheckpoint.calc_test_auc(\n",
" test_generator, hybrid_mod, test_samples=test_generator.nb_sample, return_y_res=True)\n",
"print hybrid_auc"
]
},
{
"cell_type": "code",
"execution_count": 7,
"metadata": {
"collapsed": false
},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"True\n",
"True\n"
]
}
],
"source": [
"print np.array_equal(res_y_true, vgg_y_true)\n",
"print np.array_equal(res_y_true, hybrid_y_true)"
]
},
{
"cell_type": "code",
"execution_count": 9,
"metadata": {
"collapsed": false
},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"0.88507827086\n"
]
}
],
"source": [
"from sklearn.metrics import roc_auc_score\n",
"all_mod_y_pred_avg = (res_y_pred[:,1] + vgg_y_pred[:,1] + hybrid_y_pred[:,1])/3\n",
"print roc_auc_score(res_y_true[:,1], all_mod_y_pred_avg)"
]
},
{
"cell_type": "code",
"execution_count": 10,
"metadata": {
"collapsed": false
},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"[ 0.87832831 0.87832831]\n"
]
}
],
"source": [
"from dm_keras_ext import DMAucModelCheckpoint\n",
"res_auc_aug, res_y_true_aug, res_y_pred_aug = DMAucModelCheckpoint.calc_test_auc(\n",
" test_generator, res_mod, test_samples=test_generator.nb_sample, return_y_res=True, test_augment=True)\n",
"print res_auc_aug"
]
},
{
"cell_type": "code",
"execution_count": 11,
"metadata": {
"collapsed": false
},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"[ 0.85919862 0.85919862]\n"
]
}
],
"source": [
"from dm_keras_ext import DMAucModelCheckpoint\n",
"vgg_auc_aug, vgg_y_true_aug, vgg_y_pred_aug = DMAucModelCheckpoint.calc_test_auc(\n",
" test_generator, vgg_mod, test_samples=test_generator.nb_sample, return_y_res=True, test_augment=True)\n",
"print vgg_auc_aug"
]
},
{
"cell_type": "code",
"execution_count": 12,
"metadata": {
"collapsed": false
},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"[ 0.88237829 0.88237829]\n"
]
}
],
"source": [
"from dm_keras_ext import DMAucModelCheckpoint\n",
"hybrid_auc_aug, hybrid_y_true_aug, hybrid_y_pred_aug = DMAucModelCheckpoint.calc_test_auc(\n",
" test_generator, hybrid_mod, test_samples=test_generator.nb_sample, return_y_res=True, test_augment=True)\n",
"print hybrid_auc_aug"
]
},
{
"cell_type": "code",
"execution_count": 13,
"metadata": {
"collapsed": false
},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"0.907827086026\n"
]
}
],
"source": [
"from sklearn.metrics import roc_auc_score\n",
"all_mod_y_pred_avg_aug = (res_y_pred_aug[:,1] + vgg_y_pred_aug[:,1] + hybrid_y_pred_aug[:,1])/3\n",
"print roc_auc_score(res_y_true_aug[:,1], all_mod_y_pred_avg_aug)"
]
},
{
"cell_type": "markdown",
"metadata": {
"collapsed": true
},
"source": [
"Done."
]
}
],
"metadata": {
"kernelspec": {
"display_name": "Python 2",
"language": "python",
"name": "python2"
},
"language_info": {
"codemirror_mode": {
"name": "ipython",
"version": 2
},
"file_extension": ".py",
"mimetype": "text/x-python",
"name": "python",
"nbconvert_exporter": "python",
"pygments_lexer": "ipython2",
"version": "2.7.12"
}
},
"nbformat": 4,
"nbformat_minor": 0
}

0 comments on commit 114e2ed

Please sign in to comment.