wm_from_word_to_models.ipynb

{
 "cells": [
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "# From text to models in World Modelers"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Choose text corpus to process"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "text = \"\"\"The government promotes improved cultivar to boost agricultural production for ensuring food\n",
    "security. However, the policy to seriously cut down the use of inorganic fertilizer and phase out\n",
    "the fertilizer subsidy results in deteriorating biophysical conditions, low use of inorganic\n",
    "fertilizer, less water, significantly reduced farm sizes which lead to low benefit from the\n",
    "improved cultivar.\"\"\""
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Read text via interface to Eidos system"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "collapsed": true
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "WARNING: indra/eidos - More than one attachment to event.\n",
      "WARNING: indra/eidos - More than one attachment to event.\n",
      "WARNING: indra/eidos - More than one attachment to event.\n",
      "WARNING: indra/eidos - More than one attachment to event.\n",
      "WARNING: indra/eidos - More than one attachment to event.\n",
      "WARNING: indra/eidos - More than one attachment to event.\n",
      "WARNING: indra/eidos - More than one attachment to event.\n",
      "WARNING: indra/eidos - More than one attachment to event.\n",
      "WARNING: indra/eidos - More than one attachment to event.\n",
      "WARNING: indra/eidos - More than one attachment to event.\n"
     ]
    }
   ],
   "source": [
    "from indra.sources import eidos\n",
    "eidos_processor = eidos.process_text(text)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## De-duplicated INDRA Influence Statements"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "statements = eidos_processor.statements"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 4,
   "metadata": {
    "collapsed": true
   },
   "outputs": [
    {
     "name": "stderr",
     "output_type": "stream",
     "text": [
      "INFO: indra/assemble_corpus - Combining duplicates on 18 statements...\n",
      "INFO: indra/assemble_corpus - 10 unique statements\n",
      "INFO: indra/assemble_corpus - Combining related on 10 statements...\n",
      "INFO: indra/preassembler - combine_related: poolsize not set, not using multiprocessing.\n",
      "INFO: indra/preassembler - Preassembling Influence (10)\n",
      "INFO: indra/preassembler - Groups: 10 parent, 0 worker\n",
      "INFO: indra/preassembler - Done running parent process groups\n",
      "INFO: indra/preassembler - Done.\n",
      "INFO: indra/assemble_corpus - Filtering 10 statements for top-level...\n",
      "INFO: indra/assemble_corpus - 10 statements after filter...\n",
      "INFO: indra/assemble_corpus - 10 top-level statements\n"
     ]
    }
   ],
   "source": [
    "import indra.tools.assemble_corpus as ac\n",
    "statements = ac.run_preassembly(statements)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/plain": [
       "[Influence(agricultural production(positive), food security()),\n",
       " Influence(biophysical conditions(negative), benefit()),\n",
       " Influence(cultivar(positive), agricultural production(positive)),\n",
       " Influence(cultivar(positive), benefit()),\n",
       " Influence(farm sizes(negative), benefit()),\n",
       " Influence(government(), agricultural production(positive)),\n",
       " Influence(government(), cultivar(positive)),\n",
       " Influence(policy(), biophysical conditions(negative)),\n",
       " Influence(use of inorganic fertilizer(), farm sizes(negative)),\n",
       " Influence(water(negative), farm sizes(negative))]"
      ]
     },
     "execution_count": 5,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "statements"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Assemble into graph and visualize"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "from indra.assemblers import GraphAssembler\n",
    "ga = GraphAssembler(statements)\n",
    "ga.make_model()\n",
    "ga.save_pdf('model_graph.pdf')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 7,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "text/html": [
       "\n",
       "        <iframe\n",
       "            width=\"600\"\n",
       "            height=\"500\"\n",
       "            src=\"model_graph.pdf\"\n",
       "            frameborder=\"0\"\n",
       "            allowfullscreen\n",
       "        ></iframe>\n",
       "        "
      ],
      "text/plain": [
       "<IPython.lib.display.IFrame at 0x10ecfab70>"
      ]
     },
     "execution_count": 7,
     "metadata": {},
     "output_type": "execute_result"
    }
   ],
   "source": [
    "from IPython.display import IFrame\n",
    "IFrame(\"model_graph.pdf\", width=600, height=500)"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Assemble into a system of ODEs and simulate"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "from indra.assemblers import PysbAssembler\n",
    "from pysb.simulator import ScipyOdeSimulator; import numpy; import matplotlib.pyplot as plt; plt.ion()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [],
   "source": [
    "pa = PysbAssembler()\n",
    "pa.add_statements(statements)\n",
    "model = pa.make_model()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [],
   "source": [
    "import numpy\n",
    "ts = numpy.linspace(0, 50000)\n",
    "sim = ScipyOdeSimulator(model, ts)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 11,
   "metadata": {},
   "outputs": [],
   "source": [
    "res = sim.run()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {
    "scrolled": true
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "<matplotlib.legend.Legend at 0x1436adda0>"
      ]
     },
     "execution_count": 17,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAkcAAAERCAYAAABxfAOAAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nUsb/B/FeRYPGLQvGAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x143696320>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "plt.figure(figsize=(7,3))\n",
    "plt.xlim(0, 50000)\n",
    "plt.ylim(0, 100000)\n",
    "plt.xlabel('Time')\n",
    "plt.ylabel('Relative amount (a.u.)')\n",
    "plt.xticks([])\n",
    "plt.yticks([])\n",
    "for i, cp in enumerate(model.species):\n",
    "    if str(cp) == 'benefit()':\n",
    "        plt.plot(ts, res.all['__s%d' % i], label=str(cp))\n",
    "plt.legend()"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Assemble into a Bayesian Network represented as a Figaro model"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "from indra.assemblers import FigaroAssembler"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "fa = FigaroAssembler(statements, readout='benefit')"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "fa.make_model()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "import com.cra.figaro.language._\n",
      "import com.cra.figaro.library.atomic._\n",
      "import com.cra.figaro.library.compound._\n",
      "import com.cra.figaro.library.atomic.continuous.Normal\n",
      "import com.cra.figaro.algorithm.sampling.Importance\n",
      "\n",
      "object IndraModel {\n",
      "val biophysical_conditions = Chain(policy, (v:Double) => Normal(v, 0.2))\n",
      "val benefit = Chain(^^(biophysical_conditions,cultivar,farm_sizes), (v:(Double,Double,Double)) => Normal(+ 1.00 * v._1 + 1.00 * v._2 + 1.00 * v._3, 0.2))\n",
      "val government = Normal(1,0.2)\n",
      "val cultivar = Chain(government, (v:Double) => Normal(v, 0.2))\n",
      "val policy = Normal(1,0.2)\n",
      "val agricultural_production = Chain(^^(government,cultivar), (v:(Double,Double)) => Normal(+ 1.00 * v._1 + 1.00 * v._2, 0.2))\n",
      "val use_of_inorganic_fertilizer = Normal(1,0.2)\n",
      "val farm_sizes = Chain(^^(use_of_inorganic_fertilizer,water), (v:(Double,Double)) => Normal(+ 1.00 * v._1 + 1.00 * v._2, 0.2))\n",
      "val water = Normal(1,0.2)\n",
      "val food_security = Chain(agricultural_production, (v:Double) => Normal(v, 0.2))\n",
      "def main(args:Array[String]) = {var importance = Importance(100000, benefit)\n",
      "importance.start()\n",
      "val expv = importance.computeExpectation(benefit, (v: Double) => v)\n",
      "println(expv)}\n",
      "}\n"
     ]
    }
   ],
   "source": [
    "print(fa.print_model('IndraModel.scala'))"
   ]
  }
 ],
 "metadata": {
  "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": "3.5.2"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}