{ "cells": [ { "cell_type": "markdown", "metadata": {}, "source": [ "## Active Learning\n", "\n", "The class provides tools for active learning of regression models using expected model change (EMC) and query by committee (QBC) methods and approaches for distribution shift alleviations.\n", "\n", "The central idea behind any active learning approach is to achieve lower prediction errors (higher accuracy) in machine learning models by choosing less but more informative data points.\n", "\n", "The implementation of this algorithm follows an interactive approach. In other words, we often ask you to provide labels for the selected data points.\n", "\n", "Here is a list of main methods to carry out an active learning approach:\n", "\n", "- initialize\n", "- deposit\n", "- ignore\n", "- search\n", "- visualize" ] }, { "cell_type": "code", "execution_count": 1, "metadata": {}, "outputs": [], "source": [ "import warnings\n", "warnings.filterwarnings('ignore')\n", "\n", "from chemml.optimization import ActiveLearning\n", "from chemml.datasets import load_organic_density\n", "import numpy as np\n", "import pandas as pd" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "For this tutorial, we load a sample data from ChemML datasets." ] }, { "cell_type": "code", "execution_count": 2, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "labels : density, , shape: (500, 1)\n", "features: molecular descriptors, , shape: (500, 200)\n" ] } ], "source": [ "smi, density, features = load_organic_density()\n", "print('labels : density, ', type(density), ', shape:', density.shape)\n", "print('features: molecular descriptors,', type(features), ', shape:', features.shape)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "### Keras model example\n", "\n", "The current version of the EMC approach only work with Keras deep learning models. You should create a function that returns a Keras model. You can also set the optimizer parameters and compile the model inside the function. Here is a toy example for the Keras model:" ] }, { "cell_type": "code", "execution_count": 3, "metadata": {}, "outputs": [], "source": [ "from keras.layers import Input, Dense, Concatenate\n", "from keras.models import Sequential, Model\n", "from keras.optimizers import Adam\n", "from keras.callbacks import EarlyStopping\n", "from keras.initializers import glorot_uniform\n", "from keras import regularizers\n", "from keras import losses\n", "from keras import backend as K" ] }, { "cell_type": "code", "execution_count": 4, "metadata": {}, "outputs": [], "source": [ "def model_creator(nneurons=features.shape[1], activation=['relu','tanh'], lr = 0.001):\n", " # branch 1\n", " b1_in = Input(shape=(nneurons, ), name='inp1')\n", " l1 = Dense(12, name='l1', activation=activation[0])(b1_in)\n", " b1_l1 = Dense(6, name='b1_l1', activation=activation[0])(l1)\n", " b1_l2 = Dense(3, name='b1_l2', activation=activation[0])(b1_l1)\n", " # branch 2\n", " b2_l1 = Dense(16, name='b2_l1', activation=activation[1])(l1)\n", " b2_l2 = Dense(8, name='b2_l2', activation=activation[1])(b2_l1)\n", " # merge branches\n", " merged = Concatenate(name='merged')([b1_l2, b2_l2])\n", " # linear output\n", " out = Dense(1, name='outp', activation='linear')(merged)\n", " ###\n", " model = Model(inputs = b1_in, outputs = out)\n", " adam = Adam(lr=lr, beta_1=0.9, beta_2=0.999, epsilon=1e-8, decay=0.0)\n", " model.compile(optimizer = adam,\n", " loss = 'mean_squared_error',\n", " metrics=['mean_absolute_error'])\n", " return model" ] }, { "cell_type": "code", "execution_count": 5, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "Model: \"model\"\n", "__________________________________________________________________________________________________\n", "Layer (type) Output Shape Param # Connected to \n", "==================================================================================================\n", "inp1 (InputLayer) [(None, 200)] 0 \n", "__________________________________________________________________________________________________\n", "l1 (Dense) (None, 12) 2412 inp1[0][0] \n", "__________________________________________________________________________________________________\n", "b1_l1 (Dense) (None, 6) 78 l1[0][0] \n", "__________________________________________________________________________________________________\n", "b2_l1 (Dense) (None, 16) 208 l1[0][0] \n", "__________________________________________________________________________________________________\n", "b1_l2 (Dense) (None, 3) 21 b1_l1[0][0] \n", "__________________________________________________________________________________________________\n", "b2_l2 (Dense) (None, 8) 136 b2_l1[0][0] \n", "__________________________________________________________________________________________________\n", "merged (Concatenate) (None, 11) 0 b1_l2[0][0] \n", " b2_l2[0][0] \n", "__________________________________________________________________________________________________\n", "outp (Dense) (None, 1) 12 merged[0][0] \n", "==================================================================================================\n", "Total params: 2,867\n", "Trainable params: 2,867\n", "Non-trainable params: 0\n", "__________________________________________________________________________________________________\n" ] } ], "source": [ "m = model_creator()\n", "m.summary()" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "To view workflow, you need to install `pydot`. For example with `pip install pydot`. Then run the following code" ] }, { "cell_type": "code", "execution_count": 6, "metadata": {}, "outputs": [ { "data": { "image/svg+xml": [ "\n", "\n", "G\n", "\n", "\n", "\n", "139714280881264\n", "\n", "inp1: InputLayer\n", "\n", "\n", "\n", "139711195269152\n", "\n", "l1: Dense\n", "\n", "\n", "\n", "139714280881264->139711195269152\n", "\n", "\n", "\n", "\n", "\n", "139711186458752\n", "\n", "b1_l1: Dense\n", "\n", "\n", "\n", "139711195269152->139711186458752\n", "\n", "\n", "\n", "\n", "\n", "139711186036720\n", "\n", "b2_l1: Dense\n", "\n", "\n", "\n", "139711195269152->139711186036720\n", "\n", "\n", "\n", "\n", "\n", "139711185897600\n", "\n", "b1_l2: Dense\n", "\n", "\n", "\n", "139711186458752->139711185897600\n", "\n", "\n", "\n", "\n", "\n", "139711185898656\n", "\n", "b2_l2: Dense\n", "\n", "\n", "\n", "139711186036720->139711185898656\n", "\n", "\n", "\n", "\n", "\n", "139711186054112\n", "\n", "merged: Concatenate\n", "\n", "\n", "\n", "139711185897600->139711186054112\n", "\n", "\n", "\n", "\n", "\n", "139711185898656->139711186054112\n", "\n", "\n", "\n", "\n", "\n", "139711186095552\n", "\n", "outp: Dense\n", "\n", "\n", "\n", "139711186054112->139711186095552\n", "\n", "\n", "\n", "\n", "" ], "text/plain": [ "" ] }, "execution_count": 6, "metadata": {}, "output_type": "execute_result" } ], "source": [ "from IPython.display import SVG\n", "from keras.utils.vis_utils import model_to_dot\n", "\n", "SVG(model_to_dot(m).create(prog='dot', format='svg'))" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "### Initialize active learning: EMC\n", "\n", "You first need to instantiate the active learning class using the model_creator function and the pool of candidates (i.e., U) represented by your choice of features." ] }, { "cell_type": "code", "execution_count": 7, "metadata": {}, "outputs": [], "source": [ "al = ActiveLearning(\n", " model_creator = model_creator,\n", " U = features,\n", " target_layer = ['b1_l2', 'b2_l2'], # we could also enter 'merged' layer because it only does the concatenation\n", " train_size = 50, # 50 initial training data will be selected randomly\n", " test_size = 50, # 50 independent test data will be selected randomly for the entire search\n", " batch_size = [20,0,0] # at each round of AL, labels for 20 candidates will be queried using EMC method\n", " )" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "you first need to initialize the method to get the indices for the training and test data:" ] }, { "cell_type": "code", "execution_count": 8, "metadata": {}, "outputs": [], "source": [ "tr_ind, te_ind = al.initialize()" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "The array of queried points are always available via the queries attribute" ] }, { "cell_type": "code", "execution_count": 9, "metadata": {}, "outputs": [ { "data": { "text/plain": [ "[['initial training set',\n", " array([445, 171, 373, 342, 372, 99, 331, 195, 199, 130, 140, 201, 451,\n", " 300, 294, 108, 369, 79, 328, 398, 485, 426, 43, 391, 71, 124,\n", " 495, 276, 316, 383, 498, 225, 440, 148, 122, 35, 178, 456, 253,\n", " 216, 131, 374, 419, 357, 211, 389, 458, 324, 415, 355])],\n", " ['test set',\n", " array([ 51, 48, 223, 254, 118, 258, 399, 231, 462, 292, 438, 266, 97,\n", " 28, 119, 103, 464, 242, 101, 250, 321, 468, 144, 123, 206, 227,\n", " 410, 403, 326, 229, 363, 39, 407, 416, 38, 430, 441, 465, 78,\n", " 346, 13, 281, 256, 85, 273, 151, 404, 161, 370, 431])]]" ] }, "execution_count": 9, "metadata": {}, "output_type": "execute_result" } ], "source": [ "al.queries" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Once you acquire labels for the queried data (e.g., by experiment or simulation) give them to the al object using deposit method. This can be done partially or all at once." ] }, { "cell_type": "code", "execution_count": 10, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "we stored 50 of passed indices. A list of them is in the 'last_deposited_indices_' attribute.\n", "we stored 50 of passed indices. A list of them is in the 'last_deposited_indices_' attribute.\n" ] }, { "data": { "text/plain": [ "True" ] }, "execution_count": 10, "metadata": {}, "output_type": "execute_result" } ], "source": [ "al.deposit(tr_ind, density.values.reshape(-1,1)[tr_ind])\n", "al.deposit(te_ind, density.values.reshape(-1,1)[te_ind])" ] }, { "cell_type": "code", "execution_count": 11, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "(50, 200) (50, 1)\n", "(50, 200) (50, 1)\n" ] } ], "source": [ "print (al.X_train.shape, al.Y_train.shape)\n", "print (al.X_test.shape, al.Y_test.shape)" ] }, { "cell_type": "code", "execution_count": 12, "metadata": {}, "outputs": [ { "data": { "text/plain": [ "[]" ] }, "execution_count": 12, "metadata": {}, "output_type": "execute_result" } ], "source": [ "al.queries" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "You can not initialize the method or deposit any more data once you empty the list of queries by providing requested data." ] }, { "cell_type": "code", "execution_count": 13, "metadata": {}, "outputs": [], "source": [ "#al.initialize()\n", "# the above code raises following error message:\n", "# ValueError: The class has been already initialized and it can not be initialized again!" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "### Start the active search" ] }, { "cell_type": "code", "execution_count": 14, "metadata": {}, "outputs": [], "source": [ "q = al.search(n_evaluation=1, epochs=10, verbose=0)" ] }, { "cell_type": "code", "execution_count": 15, "metadata": {}, "outputs": [ { "data": { "text/plain": [ "array([ 3, 14, 44, 76, 106, 112, 114, 129, 147, 166, 169, 237, 248,\n", " 284, 295, 310, 343, 348, 375, 408])" ] }, "execution_count": 15, "metadata": {}, "output_type": "execute_result" } ], "source": [ "q" ] }, { "cell_type": "code", "execution_count": 16, "metadata": {}, "outputs": [ { "data": { "text/plain": [ "[['batch #1',\n", " array([ 3, 14, 44, 76, 106, 112, 114, 129, 147, 166, 169, 237, 248,\n", " 284, 295, 310, 343, 348, 375, 408])]]" ] }, "execution_count": 16, "metadata": {}, "output_type": "execute_result" } ], "source": [ "al.queries" ] }, { "cell_type": "code", "execution_count": 17, "metadata": {}, "outputs": [ { "data": { "text/html": [ "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
num_querynum_trainingnum_testmaemae_stdrmsermse_stdr2r2_std
00505044.1555720.056.9862430.00.4719230.0
\n", "
" ], "text/plain": [ " num_query num_training num_test mae mae_std rmse rmse_std \\\n", "0 0 50 50 44.155572 0.0 56.986243 0.0 \n", "\n", " r2 r2_std \n", "0 0.471923 0.0 " ] }, "execution_count": 17, "metadata": {}, "output_type": "execute_result" } ], "source": [ "al.results" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "now you need to deposit the labels for the first batch of queried data using EMC approach." ] }, { "cell_type": "code", "execution_count": 18, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "we stored 20 of passed indices. A list of them is in the 'last_deposited_indices_' attribute.\n" ] }, { "data": { "text/plain": [ "True" ] }, "execution_count": 18, "metadata": {}, "output_type": "execute_result" } ], "source": [ "al.deposit(q, density.values.reshape(-1,1)[q])" ] }, { "cell_type": "code", "execution_count": 19, "metadata": {}, "outputs": [], "source": [ "q = al.search(epochs=10, verbose=0)" ] }, { "cell_type": "code", "execution_count": 20, "metadata": {}, "outputs": [ { "data": { "text/plain": [ "array([ 15, 25, 32, 110, 153, 165, 173, 190, 197, 228, 269, 304, 308,\n", " 365, 376, 402, 422, 433, 442, 494])" ] }, "execution_count": 20, "metadata": {}, "output_type": "execute_result" } ], "source": [ "q" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Here is a data frame of the results in terms of mean absolute error, root mean square error, and r-squared" ] }, { "cell_type": "code", "execution_count": 21, "metadata": {}, "outputs": [ { "data": { "text/html": [ "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
num_querynum_trainingnum_testmaemae_stdrmsermse_stdr2r2_std
00505044.1555720.00000056.9862430.0000000.4719230.000000
11705047.2305027.99068959.0147359.1355260.4200880.184212
\n", "
" ], "text/plain": [ " num_query num_training num_test mae mae_std rmse \\\n", "0 0 50 50 44.155572 0.000000 56.986243 \n", "1 1 70 50 47.230502 7.990689 59.014735 \n", "\n", " rmse_std r2 r2_std \n", "0 0.000000 0.471923 0.000000 \n", "1 9.135526 0.420088 0.184212 " ] }, "execution_count": 21, "metadata": {}, "output_type": "execute_result" } ], "source": [ "al.results" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "### Get a baseline: random sampling\n", "\n", "You can get a baseline learning curve by running a `random_search`. This method starts from the initial sets of training and test data, and adds the same number of data as batch_size every round." ] }, { "cell_type": "code", "execution_count": 22, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "WARNING:tensorflow:5 out of the last 23 calls to .predict_function at 0x7f110b915940> triggered tf.function retracing. Tracing is expensive and the excessive number of tracings could be due to (1) creating @tf.function repeatedly in a loop, (2) passing tensors with different shapes, (3) passing Python objects instead of tensors. For (1), please define your @tf.function outside of the loop. For (2), @tf.function has experimental_relax_shapes=True option that relaxes argument shapes that can avoid unnecessary retracing. For (3), please refer to https://www.tensorflow.org/guide/function#controlling_retracing and https://www.tensorflow.org/api_docs/python/tf/function for more details.\n" ] }, { "data": { "text/plain": [ "True" ] }, "execution_count": 22, "metadata": {}, "output_type": "execute_result" } ], "source": [ "al.random_search(density.values.reshape(-1,1), n_evaluation=2, epochs=10, verbose=0)" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "The results of random sampling is also accessible using `random_results` attribute:" ] }, { "cell_type": "code", "execution_count": 23, "metadata": {}, "outputs": [ { "data": { "text/html": [ "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
num_querynum_trainingnum_testmaemae_stdrmsermse_stdr2r2_std
00505037.1483705.83167147.5025529.4385520.6185770.145817
11705042.0913331.04968257.2090131.8276100.4672430.034004
\n", "
" ], "text/plain": [ " num_query num_training num_test mae mae_std rmse \\\n", "0 0 50 50 37.148370 5.831671 47.502552 \n", "1 1 70 50 42.091333 1.049682 57.209013 \n", "\n", " rmse_std r2 r2_std \n", "0 9.438552 0.618577 0.145817 \n", "1 1.827610 0.467243 0.034004 " ] }, "execution_count": 23, "metadata": {}, "output_type": "execute_result" } ], "source": [ "al.random_results" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "### Visualize distributions\n", "\n", "We keep track of many metrics to analyze your active learning search. You can store the info during your search or visualize them at each round. The `visualize` method is able to return a dictionary of matplotlib figures for the distribution of features and labels, and the learning curves." ] }, { "cell_type": "code", "execution_count": 24, "metadata": {}, "outputs": [], "source": [ "plots = al.visualize(density.values.reshape(-1,1)) # Let's provide all the labels, now that we have all of them" ] }, { "cell_type": "code", "execution_count": 25, "metadata": {}, "outputs": [ { "data": { "text/plain": [ "{'dist_pc': (
,\n", "
,\n", "
),\n", " 'dist_y': (
,\n", "
,\n", "
),\n", " 'learning_curve':
}" ] }, "execution_count": 25, "metadata": {}, "output_type": "execute_result" } ], "source": [ "plots" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "The distribution of the first principal component:" ] }, { "cell_type": "code", "execution_count": 26, "metadata": {}, "outputs": [ { "data": { "image/png": "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\n", "text/plain": [ "
" ] }, "execution_count": 26, "metadata": {}, "output_type": "execute_result" } ], "source": [ "%matplotlib inline\n", "plots['dist_pc'][0]" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "The distribution of the labels:" ] }, { "cell_type": "code", "execution_count": 27, "metadata": {}, "outputs": [ { "data": { "image/png": "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\n", "text/plain": [ "
" ] }, "execution_count": 27, "metadata": {}, "output_type": "execute_result" } ], "source": [ "plots['dist_y'][0]" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "The learning curves:" ] }, { "cell_type": "code", "execution_count": 28, "metadata": {}, "outputs": [ { "data": { "image/png": "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\n", "text/plain": [ "
" ] }, "execution_count": 28, "metadata": {}, "output_type": "execute_result" } ], "source": [ "plots['learning_curve']" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "### Search in a loop\n", "\n", "This is how we use this module in a loop:" ] }, { "cell_type": "code", "execution_count": 29, "metadata": {}, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "we stored 100 of passed indices. A list of them is in the 'last_deposited_indices_' attribute.\n", "we stored 50 of passed indices. A list of them is in the 'last_deposited_indices_' attribute.\n", "WARNING:tensorflow:6 out of the last 25 calls to .predict_function at 0x7f110b89c9d0> triggered tf.function retracing. Tracing is expensive and the excessive number of tracings could be due to (1) creating @tf.function repeatedly in a loop, (2) passing tensors with different shapes, (3) passing Python objects instead of tensors. For (1), please define your @tf.function outside of the loop. For (2), @tf.function has experimental_relax_shapes=True option that relaxes argument shapes that can avoid unnecessary retracing. For (3), please refer to https://www.tensorflow.org/guide/function#controlling_retracing and https://www.tensorflow.org/api_docs/python/tf/function for more details.\n", "we stored 50 of passed indices. A list of them is in the 'last_deposited_indices_' attribute.\n", "WARNING:tensorflow:5 out of the last 23 calls to .predict_function at 0x7f10d81f3f70> triggered tf.function retracing. Tracing is expensive and the excessive number of tracings could be due to (1) creating @tf.function repeatedly in a loop, (2) passing tensors with different shapes, (3) passing Python objects instead of tensors. For (1), please define your @tf.function outside of the loop. For (2), @tf.function has experimental_relax_shapes=True option that relaxes argument shapes that can avoid unnecessary retracing. For (3), please refer to https://www.tensorflow.org/guide/function#controlling_retracing and https://www.tensorflow.org/api_docs/python/tf/function for more details.\n", "we stored 50 of passed indices. A list of them is in the 'last_deposited_indices_' attribute.\n", "WARNING:tensorflow:5 out of the last 23 calls to .predict_function at 0x7f107c3d9ee0> triggered tf.function retracing. Tracing is expensive and the excessive number of tracings could be due to (1) creating @tf.function repeatedly in a loop, (2) passing tensors with different shapes, (3) passing Python objects instead of tensors. For (1), please define your @tf.function outside of the loop. For (2), @tf.function has experimental_relax_shapes=True option that relaxes argument shapes that can avoid unnecessary retracing. For (3), please refer to https://www.tensorflow.org/guide/function#controlling_retracing and https://www.tensorflow.org/api_docs/python/tf/function for more details.\n", "we stored 50 of passed indices. A list of them is in the 'last_deposited_indices_' attribute.\n", "WARNING:tensorflow:5 out of the last 23 calls to .predict_function at 0x7f107c6bd550> triggered tf.function retracing. Tracing is expensive and the excessive number of tracings could be due to (1) creating @tf.function repeatedly in a loop, (2) passing tensors with different shapes, (3) passing Python objects instead of tensors. For (1), please define your @tf.function outside of the loop. For (2), @tf.function has experimental_relax_shapes=True option that relaxes argument shapes that can avoid unnecessary retracing. For (3), please refer to https://www.tensorflow.org/guide/function#controlling_retracing and https://www.tensorflow.org/api_docs/python/tf/function for more details.\n", "we stored 50 of passed indices. A list of them is in the 'last_deposited_indices_' attribute.\n", "WARNING:tensorflow:5 out of the last 23 calls to .predict_function at 0x7f10d855a160> triggered tf.function retracing. Tracing is expensive and the excessive number of tracings could be due to (1) creating @tf.function repeatedly in a loop, (2) passing tensors with different shapes, (3) passing Python objects instead of tensors. For (1), please define your @tf.function outside of the loop. For (2), @tf.function has experimental_relax_shapes=True option that relaxes argument shapes that can avoid unnecessary retracing. For (3), please refer to https://www.tensorflow.org/guide/function#controlling_retracing and https://www.tensorflow.org/api_docs/python/tf/function for more details.\n", "we stored 50 of passed indices. A list of them is in the 'last_deposited_indices_' attribute.\n" ] } ], "source": [ "import os\n", "warnings.filterwarnings('ignore')\n", "\n", "out_dir = 'al_toy_example' # the arbitrary path to the output directory\n", "if not os.path.exists(out_dir):\n", " os.makedirs(out_dir)\n", "\n", "# all data sets must be 2-dimensional\n", "y = density.values.reshape(-1,1)\n", "\n", "al = ActiveLearning(\n", " model_creator = model_creator,\n", " U = features,\n", " target_layer = ['b1_l2', 'b2_l2'], # we could also enter 'merged' layer because it only does the concatenation\n", " train_size = 100, # 100 initial training data will be selected randomly\n", " test_size = 50, # 50 independent test data will be selected randomly for the entire search\n", " batch_size = [50,0,0] # at each round of AL, labels for 20 candidates will be queried using EMC method\n", " )\n", "\n", "tr_ind, te_ind = al.initialize()\n", "\n", "\n", "al.deposit(tr_ind, y[tr_ind])\n", "al.deposit(te_ind, y[te_ind])\n", "\n", "\n", "while al.query_number < 5:\n", " early_stopping = EarlyStopping(monitor='val_loss', min_delta=1e-6, patience=20, verbose=0, mode='auto')\n", " tr_ind = al.search(n_evaluation=3, ensemble='kfold', n_ensemble=3, normalize_input=True, normalize_internal=False,\n", " batch_size=32, epochs=5, verbose=0)#, callbacks=[early_stopping],validation_split=0.1)\n", "\n", " al.results.to_csv(out_dir+'/emc.csv',index=False)\n", "\n", " pd.DataFrame(al.train_indices).to_csv(out_dir+'/train_indices.csv',index=False)\n", " pd.DataFrame(al.test_indices).to_csv(out_dir+'/test_indices.csv',index=False)\n", "\n", " al.deposit(tr_ind, y[tr_ind])\n", "\n", " # you can run random search later but, I need it to be in my learning curve\n", " al.random_search(y, n_evaluation=3, random_state=13, batch_size=32, epochs=5, verbose=0)#, callbacks=[early_stopping],validation_split=0.05)\n", "\n", " al.random_results.to_csv(out_dir+'/random.csv',index=False)\n", "\n", " plots = al.visualize(y)\n", " if not os.path.exists(out_dir+\"/plots\"):\n", " os.makedirs(out_dir+\"/plots\")\n", "\n", " plots['dist_pc'][0].savefig(out_dir+\"/plots/dist_pc_0_%i.png\"%al.query_number , close = True, verbose = True)\n", " plots['dist_y'][0].savefig(out_dir+\"/plots/dist_y_0_%i.png\"%al.query_number , close = True, verbose = True)\n", " plots['learning_curve'].savefig(out_dir+\"/plots/lcurve_%i.png\"%al.query_number, close = True, verbose = True)\n" ] }, { "cell_type": "code", "execution_count": 30, "metadata": {}, "outputs": [ { "data": { "text/html": [ "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
num_querynum_trainingnum_testmaemae_stdrmsermse_stdr2r2_std
001005051.4268833.92819865.6190693.9012400.2973330.084853
111505043.73091410.22464255.31388113.1329430.4744160.255601
222005030.7961864.09138738.5630454.0488410.7555100.052260
332505035.2750991.67210942.9154251.1830240.7002810.016584
443005029.8248963.47371036.4961025.3308960.7787830.065220
\n", "
" ], "text/plain": [ " num_query num_training num_test mae mae_std rmse \\\n", "0 0 100 50 51.426883 3.928198 65.619069 \n", "1 1 150 50 43.730914 10.224642 55.313881 \n", "2 2 200 50 30.796186 4.091387 38.563045 \n", "3 3 250 50 35.275099 1.672109 42.915425 \n", "4 4 300 50 29.824896 3.473710 36.496102 \n", "\n", " rmse_std r2 r2_std \n", "0 3.901240 0.297333 0.084853 \n", "1 13.132943 0.474416 0.255601 \n", "2 4.048841 0.755510 0.052260 \n", "3 1.183024 0.700281 0.016584 \n", "4 5.330896 0.778783 0.065220 " ] }, "execution_count": 30, "metadata": {}, "output_type": "execute_result" } ], "source": [ "al.results" ] }, { "cell_type": "code", "execution_count": 31, "metadata": {}, "outputs": [ { "data": { "text/html": [ "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
num_querynum_trainingnum_testmaemae_stdrmsermse_stdr2r2_std
001005043.9520713.56905156.3815255.8372950.4775300.107432
111505040.9479337.82989551.70370311.5781280.5434900.208759
222005043.5024378.24224355.42661910.8147820.4814140.207281
332505028.2836383.29165735.4235233.1748390.7943090.035551
443005033.6463651.96567842.5930521.5392700.7046060.021051
\n", "
" ], "text/plain": [ " num_query num_training num_test mae mae_std rmse \\\n", "0 0 100 50 43.952071 3.569051 56.381525 \n", "1 1 150 50 40.947933 7.829895 51.703703 \n", "2 2 200 50 43.502437 8.242243 55.426619 \n", "3 3 250 50 28.283638 3.291657 35.423523 \n", "4 4 300 50 33.646365 1.965678 42.593052 \n", "\n", " rmse_std r2 r2_std \n", "0 5.837295 0.477530 0.107432 \n", "1 11.578128 0.543490 0.208759 \n", "2 10.814782 0.481414 0.207281 \n", "3 3.174839 0.794309 0.035551 \n", "4 1.539270 0.704606 0.021051 " ] }, "execution_count": 31, "metadata": {}, "output_type": "execute_result" } ], "source": [ "al.random_results" ] }, { "cell_type": "code", "execution_count": 32, "metadata": {}, "outputs": [ { "data": { "image/png": "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\n", "text/plain": [ "
" ] }, "execution_count": 32, "metadata": {}, "output_type": "execute_result" } ], "source": [ "plots = al.visualize(y)\n", "plots['learning_curve']" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "### Create a video\n", "\n", "If you store all the plots at each round you will be able to create a video similar to these ones:" ] }, { "cell_type": "code", "execution_count": 33, "metadata": {}, "outputs": [ { "data": { "text/html": [ "