{ "metadata": { "name": "", "signature": "sha256:8e57f7f51c102b9adcc878fe3e4380081053184e9da1e5ee4bb95b79d21ad614" }, "nbformat": 3, "nbformat_minor": 0, "worksheets": [ { "cells": [ { "cell_type": "heading", "level": 1, "metadata": {}, "source": [ "Easy comparison of experiments" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Once you have setup an experiment, you often want to try different configurations, e.g. change the interest model or try goal babbling instead of motor babbling. To help you do that, explauto provides you the [ExperimentPool](http://flowersteam.github.io/explauto/explauto.experiment.html#explauto.experiment.pool.ExperimentPool) class. This class allows you to:\n", "* create a set of experiments by combining all the given settings\n", "* run all these experiments in parallel (using the multiprocessing module)\n", "* evaluate all the run on the same testcases\n", "* compare all the results log" ] }, { "cell_type": "heading", "level": 2, "metadata": {}, "source": [ "Creation of a pool of experiments" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "As an illustration of how this class is working, we will re-do the same experiment as in the [previous tutorial](). We will compare motor and goal babbling for the simple arm environment with the high dimensional configuration." ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "First, we need to import the [ExperimentPool](http://flowersteam.github.io/explauto/explauto.experiment.html#explauto.experiment.pool.ExperimentPool) class." ] }, { "cell_type": "code", "collapsed": false, "input": [ "from explauto.experiment import ExperimentPool" ], "language": "python", "metadata": {}, "outputs": [], "prompt_number": 1 }, { "cell_type": "markdown", "metadata": {}, "source": [ "Then we instantiate it with all the different configurations we want to test as parameters." ] }, { "cell_type": "code", "collapsed": false, "input": [ "xps = ExperimentPool.from_settings_product(environments=[('simple_arm', 'high_dimensional')],\n", " babblings=['motor', 'goal'],\n", " interest_models=[('random', 'default')],\n", " sensorimotor_models=[('nearest_neighbor', 'default')],\n", " evaluate_at=[10, 20, 30, 50, 100, 200, 300, 400],\n", " same_testcases=True)" ], "language": "python", "metadata": {}, "outputs": [], "prompt_number": 2 }, { "cell_type": "markdown", "metadata": {}, "source": [ "You can note that contrary to the [from_settings](http://flowersteam.github.io/explauto/explauto.experiment.html#explauto.experiment.experiment.Experiment.from_settings) method, the [from_settings_product](http://flowersteam.github.io/explauto/explauto.experiment.html#explauto.experiment.pool.ExperimentPool.from_settings_product) you must specify environments (resp. interest and sensorimotor model) by giving a couple with the name of the chosen environment (resp. interest or sensorimotor model) and a configuration. Thus, you can easily compare the same environment with two different configurations." ] }, { "cell_type": "code", "collapsed": false, "input": [ "for i, xp_settings in enumerate(xps.settings):\n", " \n", " print \"\"\"Xp #{}:\n", " env='{self.environment}' conf='{self.environment_config}'\n", " babbling mode='{self.babbling_mode}' \n", " interest model='{self.interest_model}' conf='{self.interest_model_config}'\n", " sensorimotor model='{self.sensorimotor_model}' conf='{self.sensorimotor_model_config}'\"\"\".format(i, self=xp_settings)" ], "language": "python", "metadata": {}, "outputs": [ { "output_type": "stream", "stream": "stdout", "text": [ "Xp #0:\n", " env='simple_arm' conf='high_dimensional'\n", " babbling mode='motor' \n", " interest model='random' conf='default'\n", " sensorimotor model='nearest_neighbor' conf='default'\n", "Xp #1:\n", " env='simple_arm' conf='high_dimensional'\n", " babbling mode='goal' \n", " interest model='random' conf='default'\n", " sensorimotor model='nearest_neighbor' conf='default'\n" ] } ], "prompt_number": 3 }, { "cell_type": "markdown", "metadata": {}, "source": [ "The evaluate_at parameter is used to set the evaluation indices for **all** experiments. You can use the same_testcases parameter to choose whether all experiments will use the same testcases or not. Be aware that if you choose to use the same testcases it will be computed by the first experiment." ] }, { "cell_type": "heading", "level": 2, "metadata": {}, "source": [ "Running all experiments" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Once setup, you just need to run all experiments by using the usual [run](http://flowersteam.github.io/explauto/explauto.experiment.html#explauto.experiment.pool.ExperimentPool.run) method. This method will create a [Pool](https://docs.python.org/2/library/multiprocessing.html#module-multiprocessing.pool) of processes for each of the experiment. You can specifiy how many process will be runned in parallel (by default, the Pool will guess the number of available CPUs and will use them all)." ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Each process will create a new experiment (i.e. a new [Environment](http://flowersteam.github.io/explauto/explauto.environment.html#explauto.environment.environment.Environment) and [Agent](http://flowersteam.github.io/explauto/explauto.agent.html#explauto.agent.agent.Agent) instances), yet you have to make sure that multiple instances of your environment (resp. models) can be created." ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Let's run both experiments in parallel now! If you have a multi-processer computer, this should take about twice as long as in the previous tutorial." ] }, { "cell_type": "code", "collapsed": false, "input": [ "logs = xps.run()" ], "language": "python", "metadata": {}, "outputs": [], "prompt_number": 4 }, { "cell_type": "markdown", "metadata": {}, "source": [ "The [run](http://flowersteam.github.io/explauto/explauto.experiment.html#explauto.experiment.pool.ExperimentPool.run) method will end once all experiments in the pool are done. It returns a list of all the logs for each experiment." ] }, { "cell_type": "heading", "level": 2, "metadata": {}, "source": [ "Plotting results" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "You can then directly plot all learning curves on the same graph." ] }, { "cell_type": "code", "collapsed": false, "input": [ "%pylab inline\n", "\n", "ax = axes()\n", "\n", "for log in xps.logs:\n", " log.plot_learning_curve(ax)\n", " \n", "legend(('motor', 'goal'))" ], "language": "python", "metadata": {}, "outputs": [ { "output_type": "stream", "stream": "stdout", "text": [ "Populating the interactive namespace from numpy and matplotlib\n" ] }, { "metadata": {}, "output_type": "pyout", "prompt_number": 5, "text": [ "" ] }, { "metadata": {}, "output_type": "display_data", "png": "iVBORw0KGgoAAAANSUhEUgAAAXsAAAEXCAYAAABMCOQqAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nSsyyny4AAAAASUVORK5CYII=\n", "text": [ "" ] } ], "prompt_number": 5 }, { "cell_type": "markdown", "metadata": {}, "source": [ "To iterate on the configurations and logs, for instance to plot a same log for all experiment, you can use the following code:" ] }, { "cell_type": "code", "collapsed": false, "input": [ "colors = ('r', 'g', 'b')\n", "\n", "for i, (config, log) in enumerate(zip(xps.settings, xps.logs)): \n", " plot_index = 120 + i + 1\n", " ax_motor = subplot(plot_index)\n", " ax_motor.axis([0, 1, -1, 1])\n", " log.scatter_plot(ax_motor, (('sensori', [0, 1]), ), color=colors[i])\n", " \n", " legend([config.babbling_mode])" ], "language": "python", "metadata": {}, "outputs": [ { "metadata": {}, "output_type": "display_data", "png": "iVBORw0KGgoAAAANSUhEUgAAAX4AAAD9CAYAAAC7iRw+AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nxy8QCAQ9jP8Hg2s/lEONZQgAAAAASUVORK5CYII=\n", "text": [ "" ] } ], "prompt_number": 6 }, { "cell_type": "heading", "level": 1, "metadata": {}, "source": [ "Exercise" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Comment on the standard deviation of the first plot and think of a way to solve this problem." ] }, { "cell_type": "code", "collapsed": false, "input": [ "%load exercise_solutions/running_experiment_pool__pool.py" ], "language": "python", "metadata": {}, "outputs": [], "prompt_number": 7 } ], "metadata": {} } ] }