make_run_arg-check.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "from rnncomp.dataman import *\n",
    "\n",
    "import matplotlib.pyplot as plt\n",
    "import numpy as np\n",
    "%matplotlib inline"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "def load_dat_file(fi):\n",
    "    \"\"\"stupid hack to let `make_run_args` accept a file\n",
    "    or just a dat structure\"\"\"\n",
    "\n",
    "    if type(fi) == np.lib.npyio.NpzFile:\n",
    "        dat = fi[\"class_sig_list\"]\n",
    "    else:\n",
    "        dat = fi\n",
    "\n",
    "    cls_num = dat.shape[0]\n",
    "    sig_num = dat.shape[1]\n",
    "    dims = dat.shape[2]\n",
    "    t_steps = dat.shape[3]\n",
    "\n",
    "    return dat, cls_num, sig_num, dims, t_steps"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 12,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "def make_run_args_nengo(fi):\n",
    "    \"\"\"reshape before passing (stop organising by class) \n",
    "    and get the correct-ans and pass that too\n",
    "\n",
    "    the Nengo args need to be defined for each time step and include pauses\"\"\"\n",
    "\n",
    "    dat, cls_num, sig_num, dims, t_steps = load_dat_file(fi)\n",
    "\n",
    "    pause_size = int(PAUSE/dt)\n",
    "\n",
    "    # append zeros to the questions for pauses\n",
    "    tot_sigs = int(cls_num*sig_num)\n",
    "    zer = np.zeros((tot_sigs, dims, pause_size))\n",
    "    re_zer = dat.reshape((tot_sigs, dims, t_steps))\n",
    "    final_dat = np.concatenate((zer, re_zer), axis=2)\n",
    "    return final_dat"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 44,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(3, 1, 2, 100)\n"
     ]
    }
   ],
   "source": [
    "t_dat, t_cor = mk_cls_dataset(t_len=0.1, dims=2, n_classes=3, freq=0, class_type=\"flat\", save_res=False)[0]\n",
    "print(t_dat.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 45,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(3, 2, 110)\n"
     ]
    }
   ],
   "source": [
    "res = make_run_args_nengo(t_dat)\n",
    "print(res.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 54,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(0, 1.1)"
      ]
     },
     "execution_count": 54,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXgAAAD7CAYAAABgzo9kAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f27715fdc10>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "for ri in range(res.shape[0]):\n",
    "    for dim in range(res.shape[1]):\n",
    "        plt.plot(res[ri][dim])\n",
    "plt.ylim(0, 1.1)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 48,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": [
    "t1_dat = mk_cls_dataset(t_len=0.1, dims=1, n_classes=3, freq=0, class_type=\"flat\", save_res=False)[0]"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 49,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "(3, 1, 110)\n"
     ]
    }
   ],
   "source": [
    "res1 = make_run_args_nengo(t1_dat)\n",
    "print(res1.shape)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 50,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "data": {
      "text/plain": [
       "(0, 1.1)"
      ]
     },
     "execution_count": 50,
     "metadata": {},
     "output_type": "execute_result"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXgAAAD7CAYAAABgzo9kAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nJEmSJEmSJEmSJEmSJEmSJEla7/4fYKSrz2ZPOXYAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f2771a31610>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "for ri in range(res1.shape[0]):\n",
    "    plt.plot(res1[ri].squeeze())\n",
    "plt.ylim(0, 1.1)"
   ]
  }
 ],
 "metadata": {
  "kernelspec": {
   "display_name": "Python 2",
   "language": "python",
   "name": "python2"
  },
  "language_info": {
   "codemirror_mode": {
    "name": "ipython",
    "version": 2
   },
   "file_extension": ".py",
   "mimetype": "text/x-python",
   "name": "python",
   "nbconvert_exporter": "python",
   "pygments_lexer": "ipython2",
   "version": "2.7.6"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 0
}