examples/factor_fit_als_bcd.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "id": "31d498f6",
   "metadata": {},
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/parshakova.tanya/opt/anaconda3/envs/routing/lib/python3.9/site-packages/tqdm/auto.py:22: TqdmWarning: IProgress not found. Please update jupyter and ipywidgets. See https://ipywidgets.readthedocs.io/en/stable/user_install.html\n",
      "  from .autonotebook import tqdm as notebook_tqdm\n"
     ]
    }
   ],
   "source": [
    "import numpy as np\n",
    "\n",
    "import seaborn as sns\n",
    "import matplotlib.pyplot as plt\n",
    "\n",
    "import copy, time\n",
    "import random\n",
    "import pickle\n",
    "\n",
    "import mlrfit as mf"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "id": "4759ee57",
   "metadata": {},
   "outputs": [],
   "source": [
    "np.random.seed(1001)\n",
    "random.seed(1001)"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "573d6bef",
   "metadata": {},
   "source": [
    "#  Matrix definition"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "id": "b435696c",
   "metadata": {},
   "outputs": [],
   "source": [
    "m = 800\n",
    "n = 600\n",
    "\n",
    "\n",
    "mtype = \"dgt\"\n",
    " \n",
    "A = mf.dgt_matrix(m, n, d=3, k=5)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "id": "d188d749",
   "metadata": {},
   "outputs": [],
   "source": [
    "filename = \"%s_m%d_bcd_als\"%(mtype, m)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "id": "d29af477",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "num_levels=11\n"
     ]
    }
   ],
   "source": [
    "num_levels = int(np.ceil(np.log2(min(m,n)))+1)\n",
    "print(f\"{num_levels=}\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "id": "774d2b7e",
   "metadata": {},
   "outputs": [],
   "source": [
    "rank = num_levels*2"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "id": "8e62bfa5",
   "metadata": {},
   "outputs": [],
   "source": [
    "eps_ff = 0.005\n",
    "eps = 1e-3\n",
    "ranks = np.ones(num_levels, dtype=np.int32)*(rank // num_levels) \n",
    "refined = True\n",
    "warm_start = False"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "id": "0f4a1449",
   "metadata": {},
   "outputs": [],
   "source": [
    "info = {}\n",
    "methods = sorted(['als','bcd'])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "id": "ac3a7641",
   "metadata": {},
   "outputs": [],
   "source": [
    "PSD = False\n",
    "symm = False"
   ]
  },
  {
   "attachments": {},
   "cell_type": "markdown",
   "id": "4a89bb2e",
   "metadata": {},
   "source": [
    "# Spectral partitioning + greedy refinement"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "id": "b69cbed6",
   "metadata": {
    "scrolled": false
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "/Users/parshakova.tanya/Documents/mlr_fitting/mlrfit/utils.py:404: NumbaPerformanceWarning: np.dot() is faster on contiguous arrays, called on (array(float64, 2d, A), array(float64, 2d, A))\n",
      "  hat_A_except_level[r1:r2, c1:c2] += np.dot(B_level[r1:r2], C_level[c1:c2].T)\n"
     ]
    },
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "* level=0, losses[0]=0.906, losses[-1]=0.906, len(losses)=2, [2]\n",
      "* level=1, losses[0]=0.779, losses[-1]=0.779, len(losses)=2, [2 2]\n",
      "* level=2, losses[0]=0.660, losses[-1]=0.657, len(losses)=2, [2 2 2]\n",
      "* level=3, losses[0]=0.550, losses[-1]=0.546, len(losses)=2, [2 2 2 2]\n",
      "* level=4, losses[0]=0.475, losses[-1]=0.471, len(losses)=2, [2 2 2 2 2]\n",
      "* level=5, losses[0]=0.440, losses[-1]=0.438, len(losses)=2, [2 2 2 2 2 2]\n",
      "* level=6, losses[0]=0.428, losses[-1]=0.427, len(losses)=2, [2 2 2 2 2 2 2]\n",
      "* level=7, losses[0]=0.424, losses[-1]=0.423, len(losses)=2, [2 2 2 2 2 2 2 2]\n",
      "* level=8, losses[0]=0.421, losses[-1]=0.420, len(losses)=2, [2 2 2 2 2 2 2 2 2]\n",
      "* level=9, losses[0]=0.419, losses[-1]=0.418, len(losses)=2, [2 2 2 2 2 2 2 2 2 2]\n",
      "* level=10, losses[0]=0.417, losses[-1]=0.417, len(losses)=2, [2 2 2 2 2 2 2 2 2 2 2]\n",
      "level=10, loss=0.41667705544256606\n"
     ]
    }
   ],
   "source": [
    "hat_A = mf.MLRMatrix()\n",
    "td_losses, _ = hat_A.hpartition_topdown(A, ranks, eps_ff=eps_ff,\\\n",
    "                                        symm=symm, max_iters_ff=1, PSD=PSD)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "id": "c27522e6",
   "metadata": {},
   "outputs": [],
   "source": [
    "epoch_len = 10\n",
    "init_type = 'bcd'"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "id": "828b1af9",
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "itr=0, 0.49117147370330216, [2 2 2 2 2 2 2 2 2 2 2]\n",
      "itr=5, 0.42958847092065183, [2 2 2 2 2 2 2 2 2 2 2], time_epoch=1.2541661262512207\n",
      "itr=10, 0.4216396169706212, [2 2 2 2 2 2 2 2 2 2 2], time_epoch=1.35636305809021\n",
      "itr=15, 0.41592710340925765, [2 2 2 2 2 2 2 2 2 2 2], time_epoch=1.3144209384918213\n",
      "als         , m_time=26.801811933517456, losses[0]=0.49117147370330216, losses[-1]=0.4138551601413846\n",
      "itr=0, 0.49117147370330216, [2 2 2 2 2 2 2 2 2 2 2]\n",
      "itr=5, 0.42721068984232974, [2 2 2 2 2 2 2 2 2 2 2], time_v_epoch=0.8837540149688721, time_loss=0.010970830917358398\n",
      "itr=10, 0.4206363574240819, [2 2 2 2 2 2 2 2 2 2 2], time_v_epoch=0.8581357002258301, time_loss=0.008408069610595703\n",
      "itr=15, 0.41565776457850506, [2 2 2 2 2 2 2 2 2 2 2], time_v_epoch=0.8747749328613281, time_loss=0.009053945541381836\n",
      "bcd         , m_time=17.157546997070312, losses[0]=0.49117147370330216, losses[-1]=0.41360943375188747\n"
     ]
    }
   ],
   "source": [
    "for method in  methods:\n",
    "    start_time = time.time()\n",
    "    if method in ['bcd']: eps_ff1 = eps_ff/2.5\n",
    "    else: eps_ff1 = eps_ff\n",
    "    losses = hat_A.factor_fit(A, ranks, hat_A.hpart, method=method, eps_ff=eps_ff1, PSD=PSD, \\\n",
    "                                freq=5, svds_v0=True, epoch_len=epoch_len, init_type=init_type,\\\n",
    "                                printing=True, max_iters_ff=10**3, symm=symm, warm_start=False)\n",
    "    m_time = time.time()-start_time\n",
    "    \n",
    "    info[method] = {'loss':losses, 'time':m_time}\n",
    "    print(f\"{method:12s}, {m_time=}, {losses[0]=}, {losses[-1]=}\")"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 13,
   "id": "1ca828eb",
   "metadata": {},
   "outputs": [],
   "source": [
    "with open('outputs/ff_info_%s.pickle'%filename, 'wb') as handle:\n",
    "    pickle.dump(info, handle, protocol=pickle.HIGHEST_PROTOCOL)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "id": "288f6bf0",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "als               , num_iters= 18, loss=0.41385516, t=26.801811933517456\n",
      "bcd               , num_iters= 19, loss=0.41360943, t=17.157546997070312\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 600x480 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "mf.plot_ff_loss_methods(info, iterations=True, dpi=120, figsize=(5, 4), filename=\"plots/ff_itr%d_%s.pdf\"%(1, filename))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "id": "506eb8a9",
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "als               , num_iters= 18, loss=0.41385516, t=26.801811933517456\n",
      "bcd               , num_iters= 19, loss=0.41360943, t=17.157546997070312\n"
     ]
    },
    {
     "data": {
      "image/png": "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",
      "text/plain": [
       "<Figure size 600x480 with 2 Axes>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "mf.plot_ff_loss_methods(info, iterations=False, dpi=120, figsize=(5, 4), filename=\"plots/ff_itr%d_%s.pdf\"%(0, filename))"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "id": "5068b4e4",
   "metadata": {},
   "outputs": [],
   "source": []
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 3 (ipykernel)",
   "language": "python",
   "name": "python3"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 3
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython3",
   "version": "3.9.16"
  },
  "vscode": {
   "interpreter": {
    "hash": "cbb24b48ae2642bb0fe3c3a73dd180d2cf6b4e8df7de3f3850c06f8dd008ecd7"
   }
  }
 },
 "nbformat": 4,
 "nbformat_minor": 5
}