{ "cells": [ { "cell_type": "markdown", "metadata": {}, "source": [ "# Using R via rpy2" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "This example illustrates how to use models, summary statistics and\n", "distance functions defined in R. We're assuming you're already\n", "familiar with the basic workings of pyABC. If not, consult the\n", "other tutorial examples." ] }, { "cell_type": "raw", "metadata": { "raw_mimetype": "text/restructuredtext" }, "source": [ "Download this notebook :download:`Using R with pyABC `." ] }, { "cell_type": "raw", "metadata": { "raw_mimetype": "text/restructuredtext" }, "source": [ "In this example, we're introducing the new class\n", ":class:`pyabc.external.r.R ` which is our interface with R.\n", "We use this class to to load an external R script." ] }, { "cell_type": "code", "execution_count": 1, "metadata": {}, "outputs": [], "source": [ "# install if not done yet\n", "!pip install pyabc[r] --quiet" ] }, { "cell_type": "code", "execution_count": 1, "metadata": {}, "outputs": [ { "name": "stderr", "output_type": "stream", "text": [ "/home/yannik/pyabc/pyabc/external/r/r_rpy2.py:81: UserWarning: The support of R via rpy2 is considered experimental.\n", " warnings.warn(\"The support of R via rpy2 is considered experimental.\")\n" ] } ], "source": [ "%matplotlib inline\n", "from pyabc.external.r import R\n", "\n", "r = R(\"myRModel.R\")" ] }, { "cell_type": "raw", "metadata": { "raw_mimetype": "text/restructuredtext" }, "source": [ ".. note::\n", "\n", " ``R(\"myRModel.R\")`` does, amongst other things, the equivalent\n", " to R's ``source(\"myRModel.R\")``. That is, the entire script is\n", " executed with all the possible side effects this might have.\n", " \n", " \n", "You can download the file here: :download:`myRModel.R `.\n", "But now, let's have a look at it." ] }, { "cell_type": "code", "execution_count": 2, "metadata": {}, "outputs": [ { "data": { "text/html": [ "
# Blue print for pyABC parameter inference runs.\n",
       "# A proposal of how to structure an R file for usage with pyABC\n",
       "\n",
       "\n",
       "#' The model to be simulated.\n",
       "#' In this example, it is just a multivariate normal\n",
       "#' distribution. The number of parameters depends on the\n",
       "#' model and can be arbitrary. However, the parameters\n",
       "#' should be real numbers.\n",
       "#' The return type is arbitrary as well.\n",
       "myModel <- function(pars){\n",
       "  x <- rnorm(1) + pars$meanX\n",
       "  y <- rnorm(1) + pars$meanY\n",
       "  c(x,y)  # It is not important that it is a vector.\n",
       "}\n",
       "\n",
       "#' Calculates summary statistics from whatever the model returns\n",
       "#'\n",
       "#' It is important that the summary statistics have names\n",
       "#' to store them correctly in pyABC's database.\n",
       "#' In many cases, the summary statistics function might just\n",
       "#' pass through the result of the model function if the\n",
       "#' summary statistics calculation is already\n",
       "#' done there. Splitting summary statistics and the model\n",
       "#' makes most sense in a model selection scenario.\n",
       "#'\n",
       "#' @param modelResult The data simulated by the model\n",
       "#' @return Named list of summary statistics.\n",
       "mySummaryStatistics <- function(modelResult){\n",
       "  list(x=modelResult[1],\n",
       "       y=modelResult[2],\n",
       "       mtcars=mtcars,  # Can also pass data frames\n",
       "       cars=cars,\n",
       "       arbitraryKey="Some random text")\n",
       "}\n",
       "\n",
       "#' Calculate distance between summary statistics\n",
       "#'\n",
       "#' @param sumStatSample The summary statistics of the sample\n",
       "#' proposed py pyABC\n",
       "#' @param sumStatData The summary statistics of the observed\n",
       "#'        data for which we want to calculate the posterior.\n",
       "#' @return A single float\n",
       "myDistance <- function(sumStatSample, sumStatData){\n",
       "  sqrt((sumStatSample$x - sumStatData$x)^2\n",
       "       + abs(sumStatSample$y - sumStatData$y)^2)\n",
       "}\n",
       "\n",
       "\n",
       "# We store the observed data as named list\n",
       "# in a variable.\n",
       "# This is passed by pyABC as to myDistance\n",
       "# as the sumStatData argument\n",
       "mySumStatData <- list(x=4, y=8, mtcars=mtcars, cars=cars)\n",
       "\n",
       "# The functions for the model, the summary\n",
       "# statistics and distance\n",
       "# have to be constructed in\n",
       "# such a way that the following always makes sense:\n",
       "#\n",
       "# myDistance(\n",
       "#   mySummaryStatistics(myModel(list(meanX=1, meanY=2))),\n",
       "#   mySummaryStatistics(myModel(list(meanX=2, meanY=2))))\n",
       "
\n" ], "text/plain": [ "" ] }, "execution_count": 2, "metadata": {}, "output_type": "execute_result" } ], "source": [ "r.display_source_ipython()" ] }, { "cell_type": "markdown", "metadata": { "raw_mimetype": "text/restructuredtext" }, "source": [ "We see that four relevant objects are defined in the file.\n", "\n", "* myModel\n", "* mySummaryStatistics (optional)\n", "* myDistance\n", "* mySumStatData\n", "\n", "The names of these do not matter. The ``mySummaryStatistics`` is actually optional and can be omitted\n", "in case the model calculates the summary statistics directly. We load the defined functions using the ``r`` object:" ] }, { "cell_type": "code", "execution_count": 3, "metadata": {}, "outputs": [], "source": [ "model = r.model(\"myModel\")\n", "distance = r.distance(\"myDistance\")\n", "sum_stat = r.summary_statistics(\"mySummaryStatistics\")" ] }, { "cell_type": "markdown", "metadata": { "raw_mimetype": "text/restructuredtext" }, "source": [ "From there on, we can use them (almost) as if they were ordinary Python functions." ] }, { "cell_type": "code", "execution_count": 4, "metadata": {}, "outputs": [ { "name": "stderr", "output_type": "stream", "text": [ "ABC.Sampler INFO: Parallelize sampling on 4 processes.\n" ] } ], "source": [ "import pyabc\n", "from pyabc import ABCSMC, RV, Distribution\n", "\n", "pyabc.settings.set_figure_params('pyabc') # for beautified plots\n", "\n", "prior = Distribution(meanX=RV(\"uniform\", 0, 10), meanY=RV(\"uniform\", 0, 10))\n", "abc = ABCSMC(model, prior, distance, summary_statistics=sum_stat)" ] }, { "cell_type": "markdown", "metadata": { "raw_mimetype": "text/restructuredtext" }, "source": [ "We also load the observation with ``r.observation`` and pass it to a new ABC run." ] }, { "cell_type": "code", "execution_count": 5, "metadata": {}, "outputs": [ { "name": "stderr", "output_type": "stream", "text": [ "ABC.History INFO: Start \n" ] }, { "data": { "text/plain": [ "" ] }, "execution_count": 5, "metadata": {}, "output_type": "execute_result" } ], "source": [ "import os\n", "from tempfile import gettempdir\n", "\n", "db = \"sqlite:///\" + os.path.join(gettempdir(), \"test.db\")\n", "abc.new(db, r.observation(\"mySumStatData\"))" ] }, { "cell_type": "markdown", "metadata": { "raw_mimetype": "text/restructuredtext" }, "source": [ "We start a run which terminates as soon as an acceptance threshold of 0.9 or less is reached\n", "or the maximum number of 4 populations is sampled." ] }, { "cell_type": "code", "execution_count": 6, "metadata": {}, "outputs": [ { "name": "stderr", "output_type": "stream", "text": [ "ABC INFO: Calibration sample t = -1.\n", "ABC INFO: t: 0, eps: 4.33680673e+00.\n", "ABC INFO: Accepted: 100 / 240 = 4.1667e-01, ESS: 1.0000e+02.\n", "ABC INFO: t: 1, eps: 2.94456492e+00.\n", "ABC INFO: Accepted: 100 / 256 = 3.9062e-01, ESS: 9.7804e+01.\n", "ABC INFO: t: 2, eps: 1.88636339e+00.\n", "ABC INFO: Accepted: 100 / 289 = 3.4602e-01, ESS: 8.9331e+01.\n", "ABC INFO: t: 3, eps: 1.23695847e+00.\n", "ABC INFO: Accepted: 100 / 471 = 2.1231e-01, ESS: 8.4991e+01.\n", "ABC INFO: Stop: Maximum number of generations.\n", "ABC.History INFO: Done \n" ] } ], "source": [ "history = abc.run(minimum_epsilon=0.9, max_nr_populations=4)" ] }, { "cell_type": "markdown", "metadata": { "raw_mimetype": "text/restructuredtext" }, "source": [ "Lastly, we plot the results and observe how the generations contract slowly around the oserved value.\n", "(Note, that the contraction around the observed value is a particular property of the chosen example and not always the case.)" ] }, { "cell_type": "code", "execution_count": 7, "metadata": {}, "outputs": [ { "data": { "image/png": "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\n", "text/plain": [ "
" ] }, "metadata": { "needs_background": "light" }, "output_type": "display_data" }, { "data": { "image/png": "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\n", "text/plain": [ "
" ] }, "metadata": { "needs_background": "light" }, "output_type": "display_data" }, { "data": { "image/png": "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\n", "text/plain": [ "
" ] }, "metadata": { "needs_background": "light" }, "output_type": "display_data" }, { "data": { "image/png": "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\n", "text/plain": [ "
" ] }, "metadata": { "needs_background": "light" }, "output_type": "display_data" } ], "source": [ "from pyabc.visualization import plot_kde_2d\n", "\n", "for t in range(history.n_populations):\n", " df, w = abc.history.get_distribution(0, t)\n", " ax = plot_kde_2d(\n", " df,\n", " w,\n", " \"meanX\",\n", " \"meanY\",\n", " xmin=0,\n", " xmax=10,\n", " ymin=0,\n", " ymax=10,\n", " numx=100,\n", " numy=100,\n", " )\n", " ax.scatter(\n", " [4], [8], edgecolor=\"black\", facecolor=\"white\", label=\"Observation\"\n", " )\n", " ax.legend()\n", " ax.set_title(f\"PDF t={t}\")" ] }, { "cell_type": "markdown", "metadata": { "raw_mimetype": "text/restructuredtext" }, "source": [ "And we can also retrieve summary statistics such as a stored\n", "DataFrame, although the DataFrame was acutally defined in R." ] }, { "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", "
speeddist
14.02.0
24.010.0
37.04.0
47.022.0
58.016.0
\n", "
" ], "text/plain": [ " speed dist\n", "1 4.0 2.0\n", "2 4.0 10.0\n", "3 7.0 4.0\n", "4 7.0 22.0\n", "5 8.0 16.0" ] }, "execution_count": 8, "metadata": {}, "output_type": "execute_result" } ], "source": [ "history.get_weighted_sum_stats_for_model(m=0, t=1)[1][0][\"cars\"].head()" ] }, { "cell_type": "raw", "metadata": { "raw_mimetype": "text/restructuredtext" }, "source": [ "Dumping the results to a file format R can read\n", "-----------------------------------------------\n", "\n", "Although you could query pyABC's database directly from R since the database\n", "is just a SQL database (e.g. SQLite), pyABC ships with a utility for facilitate\n", "export of the database.\n", "Use the ``abc-dump`` utility provided by pyABC to dump results to\n", "file formats such as csv, feather, html, json and others.\n", "These can be easiliy read in by R. See `Exporting pyABC's database <../datastore.rst>`_ for how to use this\n", "utility.\n", "\n", "Assume you dumped to the feather format::\n", "\n", " abc-export --db results.db --out exported.feather --format feather\n", " \n", "You could read the results in with the following R snippet\n", "\n", ".. code:: R\n", "\n", "\n", " install.packages(\"feather\")\n", " install.packages(\"jsonlite\")\n", "\n", " library(\"feather\")\n", " library(\"jsonlite\")\n", "\n", " loadedDf <- data.frame(feather(\"exported.feather\"))\n", "\n", " jsonStr <- loadedDf$sumstat_ss_df[1]\n", "\n", " sumStatDf <- fromJSON(jsonStr)\n", "\n", "If you prefer CSV over the feather format you can also do that." ] } ], "metadata": { "celltoolbar": "Raw Cell Format", "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": "3.9.7" } }, "nbformat": 4, "nbformat_minor": 4 }