{ "cells": [ { "cell_type": "markdown", "metadata": {}, "source": [ "Handling TensorFlow graphs and sessions\n", "--\n", "\n", "##### Check [How GPflow relates to TensorFlow: tips & tricks](../tips_and_tricks.ipynb) for more examples and information." ] }, { "cell_type": "code", "execution_count": 4, "metadata": {}, "outputs": [], "source": [ "import gpflow\n", "import tensorflow as tf" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "\n", "TensorFlow gives you a default graph, which you fill with tensors and operations - nodes and edges in the graph respectively. See the [TensorFlow documentation on building a `tf.Graph`](https://www.tensorflow.org/guide/graphs#building_a_tfgraph) for details on how to change the default graph or exploit multiple graphs.\n", "\n", "A TensorFlow graph is a representation of your computation. To execute your code you need a [TensorFlow session](https://www.tensorflow.org/guide/graphs#executing_a_graph_in_a_tfsession). For example, you can think of graphs and sessions as binary sources and actual command running it in a terminal. Normally, TensorFlow doesn't provide the default session; however, GPflow creates a default session that you can access as shown in the following example:" ] }, { "cell_type": "code", "execution_count": 5, "metadata": {}, "outputs": [], "source": [ "session = gpflow.get_default_session()" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "To change GPflow's default session:" ] }, { "cell_type": "code", "execution_count": 6, "metadata": {}, "outputs": [], "source": [ "gpflow.reset_default_session()\n", "assert session is not gpflow.get_default_session()" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "You can manipulate sessions manually, but you have to make them the default for GPflow:" ] }, { "cell_type": "code", "execution_count": 7, "metadata": {}, "outputs": [ { "name": "stderr", "output_type": "stream", "text": [ "W0816 11:52:23.051765 140208876939072 deprecation_wrapper.py:119] From /home/alexandra/virtual_environments/prowler_venv/lib/python3.6/site-packages/gpflow/core/node.py:109: The name tf.get_default_graph is deprecated. Please use tf.compat.v1.get_default_graph instead.\n", "\n", "W0816 11:52:23.058769 140208876939072 deprecation_wrapper.py:119] From /home/alexandra/virtual_environments/prowler_venv/lib/python3.6/site-packages/gpflow/params/parameter.py:388: The name tf.placeholder is deprecated. Please use tf.compat.v1.placeholder instead.\n", "\n" ] } ], "source": [ "with tf.Session() as session:\n", " k = gpflow.kernels.SquaredExponential(input_dim=1)\n", " k.lengthscales = 2.0\n", " k.variance = 3.0" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Now, TensorFlow variables and tensors for the created `SquaredExponential` kernel are initialised with the session that was closed when the Python context ended. You can re-use the `SquaredExponential` object by re-initialising it." ] }, { "cell_type": "code", "execution_count": 8, "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", "
classpriortransformtrainableshapefixed_shapevalue
SquaredExponential/lengthscalesParameterNone+veTrue()True2.0
SquaredExponential/varianceParameterNone+veTrue()True3.0
\n", "
" ], "text/plain": [ "" ] }, "execution_count": 8, "metadata": {}, "output_type": "execute_result" } ], "source": [ "k.initialize(gpflow.get_default_session())\n", "k" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "## Using feed dicts with autoflow\n", "\n", "We'll recreate something like Figure 5.5 from [Gaussian Processes for Machine Learning](http://www.gaussianprocess.org/gpml/chapters/RW5.pdf), which is a plot of the marginal likelihood against hyperparameter configurations. We want to loop over a grid of hyperparameters, but GPflow can be very slow to do this using the assign method. We can get around this by using a feed dict with autoflow. \n" ] }, { "cell_type": "code", "execution_count": 9, "metadata": { "ExecuteTime": { "end_time": "2018-06-20T11:40:21.510889Z", "start_time": "2018-06-20T11:40:21.349272Z" } }, "outputs": [], "source": [ "import numpy as np\n", "X = np.array((-6, -4, -2.1, -2, 2, 5.5, 6)).reshape(-1, 1)\n", "Y = np.array((-0.5, -0.5, 1.7, 1.6, 1, 2, 1.9)).reshape(-1, 1)\n", "\n", "model = gpflow.models.GPR(X, Y, gpflow.kernels.RBF(1))" ] }, { "cell_type": "code", "execution_count": 10, "metadata": { "ExecuteTime": { "end_time": "2018-06-20T11:40:35.875839Z", "start_time": "2018-06-20T11:40:21.512123Z" } }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "1.67 ms ± 252 µs per loop (mean ± std. dev. of 7 runs, 100 loops each)\n" ] } ], "source": [ "%%timeit\n", "# this is how long it takes to do the calculation with fixed parameters\n", "model.compute_log_likelihood()\n", "pass" ] }, { "cell_type": "code", "execution_count": 11, "metadata": { "ExecuteTime": { "end_time": "2018-06-20T11:40:43.636125Z", "start_time": "2018-06-20T11:40:35.877115Z" } }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "2.08 ms ± 88.5 µs per loop (mean ± std. dev. of 7 runs, 100 loops each)\n" ] } ], "source": [ "%%timeit\n", "# the assign method\n", "model.kern.lengthscales = 1.\n", "model.likelihood.variance = 0.1\n", "model.compute_log_likelihood()" ] }, { "cell_type": "code", "execution_count": 9, "metadata": { "ExecuteTime": { "end_time": "2018-06-20T11:40:43.640682Z", "start_time": "2018-06-20T11:40:43.637867Z" } }, "outputs": [], "source": [ "def make_feed(param, value):\n", " return {param.unconstrained_tensor : param.transform.backward(value)}" ] }, { "cell_type": "code", "execution_count": 10, "metadata": { "ExecuteTime": { "end_time": "2018-06-20T11:40:56.972382Z", "start_time": "2018-06-20T11:40:43.642145Z" } }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "1.07 ms ± 13.1 µs per loop (mean ± std. dev. of 7 runs, 1000 loops each)\n" ] } ], "source": [ "%%timeit\n", "# the feed dict method\n", "feed_dict = {}\n", "feed_dict.update(make_feed(model.kern.lengthscales, 1.))\n", "feed_dict.update(make_feed(model.likelihood.variance, 0.1))\n", "model.compute_log_likelihood(feed_dict=feed_dict)" ] }, { "cell_type": "code", "execution_count": 11, "metadata": { "ExecuteTime": { "end_time": "2018-06-20T11:40:56.977340Z", "start_time": "2018-06-20T11:40:56.973737Z" } }, "outputs": [], "source": [ "def evaluate(func, param_feed_dict):\n", " tensor_feed_dict = {}\n", " for param, value in param_feed_dict.items():\n", " tensor_feed_dict.update(make_feed(param, value))\n", " return func(feed_dict=tensor_feed_dict)" ] }, { "cell_type": "code", "execution_count": 12, "metadata": { "ExecuteTime": { "end_time": "2018-06-20T11:40:58.371053Z", "start_time": "2018-06-20T11:40:56.978641Z" } }, "outputs": [ { "name": "stdout", "output_type": "stream", "text": [ "1.08 ms ± 22.5 µs per loop (mean ± std. dev. of 7 runs, 1000 loops each)\n" ] } ], "source": [ "%%timeit\n", "evaluate(model.compute_log_likelihood, {model.kern.lengthscales: 1., model.likelihood.variance: 0.1})" ] }, { "cell_type": "code", "execution_count": 13, "metadata": { "ExecuteTime": { "end_time": "2018-06-20T11:41:02.717145Z", "start_time": "2018-06-20T11:40:58.372035Z" } }, "outputs": [ { "data": { "image/png": "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\n", "text/plain": [ "
" ] }, "metadata": { "needs_background": "light" }, "output_type": "display_data" } ], "source": [ "import matplotlib.pyplot as plt\n", "%matplotlib inline\n", "\n", "# with the feed dict method this example is feasible\n", "\n", "log_noises = np.linspace(np.log(0.001), np.log(2.), 50)\n", "log_lengthscales = np.linspace(np.log(0.1), np.log(20), 50)\n", "\n", "xx, yy = np.meshgrid(log_lengthscales, log_noises)\n", "zz = []\n", "\n", "for lengthscale, noise in zip(np.exp(xx.flatten()), np.exp(yy.flatten())):\n", " feed_dict = {}\n", " feed_dict.update(make_feed(model.kern.lengthscales, lengthscale))\n", " feed_dict.update(make_feed(model.likelihood.variance, noise))\n", " # adapt GPflow so that we can use the following instead:\n", " #zz.append(model.evaluate(model.compute_log_likelihood,\n", " # {model.kern.lengthscales: lengthscale,\n", " # model.likelihood.variance: noise}\n", " # ))\n", " # to pass arguments e.g. for predict_f(XX), could use\n", " # model.evaluate(functools.partial(model.predict_f, XX), {...: ..., ...})\n", " zz.append(model.compute_log_likelihood(feed_dict=feed_dict))\n", " \n", "plt.contour(xx, yy, np.array(zz).reshape(xx.shape),\n", " levels = np.linspace(np.max(zz)-4, np.max(zz), 25))\n", "plt.colorbar()\n", "plt.show()" ] } ], "metadata": { "anaconda-cloud": {}, "kernelspec": { "display_name": "Python 3", "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": "3.6.8" } }, "nbformat": 4, "nbformat_minor": 2 }