theta_avg.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {},
   "outputs": [],
   "source": [
    "#%source bin/activate\n",
    "import os\n",
    "import matplotlib.pyplot as plt\n",
    "import numpy as np\n",
    "%matplotlib inline\n",
    "\n",
    "cos, sin = np.cos, np.sin\n",
    "twopi = np.pi*2.0"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "rundir = \"/home/walterms/project/walterms/mcmd/nn/tffnn/data/mixed/\"\n",
    "trndir = rundir+\"train/\"\n",
    "testdir = rundir+\"test/\"\n",
    "\n",
    "trnlist = [trndir+\"djam\", trndir+\"tjam\", trndir+\"xjam\", trndir+\"ljam\", trndir+\"ujam\"]\n",
    "testlist = [testdir+\"djam\", testdir+\"tjam\", testdir+\"xjam\", testdir+\"ljam\", testdir+\"ujam\"]\n",
    "\n",
    "fulllist = trnlist+testlist\n",
    "\n",
    "file_append = \"_th\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Processing /home/walterms/project/walterms/mcmd/nn/tffnn/data/mixed/train/djam as training data\n",
      "Processing /home/walterms/project/walterms/mcmd/nn/tffnn/data/mixed/train/tjam as training data\n",
      "Processing /home/walterms/project/walterms/mcmd/nn/tffnn/data/mixed/train/xjam as training data\n",
      "Processing /home/walterms/project/walterms/mcmd/nn/tffnn/data/mixed/train/ljam as training data\n",
      "Processing /home/walterms/project/walterms/mcmd/nn/tffnn/data/mixed/train/ujam as training data\n",
      "Processing /home/walterms/project/walterms/mcmd/nn/tffnn/data/mixed/test/djam as training data\n",
      "Processing /home/walterms/project/walterms/mcmd/nn/tffnn/data/mixed/test/tjam as training data\n",
      "Processing /home/walterms/project/walterms/mcmd/nn/tffnn/data/mixed/test/xjam as training data\n",
      "Processing /home/walterms/project/walterms/mcmd/nn/tffnn/data/mixed/test/ljam as training data\n",
      "Processing /home/walterms/project/walterms/mcmd/nn/tffnn/data/mixed/test/ujam as training data\n",
      "Done compiling training set\n"
     ]
    }
   ],
   "source": [
    "# Compile training set\n",
    "gridx, gridy = 10,10 # num cells\n",
    "\n",
    "# Edgelength is always 1.\n",
    "width,height = 1.,1.\n",
    "\n",
    "flist = fulllist\n",
    "for fname in flist:\n",
    "#     fname = flist[f]\n",
    "    print \"Processing \" + fname + \" as training data\"\n",
    "    fin = open(fname, 'r')\n",
    "    fout = open(fname+file_append, 'w')\n",
    "    thgrid = np.zeros(shape=(gridx,gridy))\n",
    "    thgrid_refs = np.zeros(shape=(gridx,gridy))\n",
    "    counts = np.zeros(shape=(gridx,gridy))\n",
    "    dx,dy = width/float(gridx), height/float(gridy)\n",
    "    nbl = 0\n",
    "    xs,ys,thetas = [],[],[]\n",
    "    specials = []\n",
    "    for line in fin.readlines():\n",
    "        if line == \"\\n\": continue\n",
    "        if line.startswith(\"label\"):\n",
    "            # Done a block\n",
    "            for x,y,th in zip(xs,ys,thetas):\n",
    "                xi = int(x/dx)\n",
    "                yi = int(y/dy)\n",
    "                i = gridx*xi + yi\n",
    "                # th is initially in range [0,1.] corresponding to [0,2pi]\n",
    "                # arctan2 outputs in range [-pi,pi]\n",
    "                # difference between arctan and arctan2 is that\n",
    "                # arctan only \"folds back\" the range [-pi/2,pi/2]\n",
    "                if th > 0.5: th = -1. + th # th: [-0.5,0.5]\n",
    "                if thgrid_refs[xi,yi] == 0.:\n",
    "                    # get it a 'safe' distance from modpoint for\n",
    "                    # angle comparisons later\n",
    "                    if th > 0.25: th -= 0.5\n",
    "                    if th < -0.25: th += 0.5\n",
    "                    if i in specials:\n",
    "                        print \"adding \", str(th)\n",
    "                        \n",
    "                    thgrid_refs[xi,yi] = th\n",
    "                else:\n",
    "                    thref = thgrid_refs[xi,yi]\n",
    "                    if (th > thref) and ((th - thref) > 0.25): th -= 0.5\n",
    "                    if (th < thref) and ((thref - th) > 0.25): th += 0.5\n",
    "\n",
    "                # return theta to [0,1] range\n",
    "                # That way empty cells will be 0. and\n",
    "                # look dead to the NN\n",
    "                th += 0.5\n",
    "                \n",
    "                thgrid[xi,yi] += th\n",
    "                counts[xi,yi] += 1\n",
    "                if i in specials:\n",
    "                    print nbl,i,th,thgrid[xi,yi]/counts[xi,yi]\n",
    "                \n",
    "            np.divide(thgrid,counts,out=thgrid,where=(counts>0.))\n",
    "            # Write block\n",
    "            dline = []\n",
    "            for i in range(gridx):\n",
    "                for j in range(gridy):\n",
    "                    dline.append(thgrid[i,j])\n",
    "            strdline = \"\"\n",
    "            for d in dline:\n",
    "                strdline += str(d)+\" \"\n",
    "            strdline = strdline[:-1]\n",
    "            fout.write(strdline+\"\\n\")\n",
    "\n",
    "            # After writing the theta grid block, append the label\n",
    "            fout.write(line+\"\\n\")\n",
    "            # Reset arrays\n",
    "            thgrid = np.zeros(shape=(gridx,gridy))\n",
    "            thgrid_refs = np.zeros(shape=(gridx,gridy))\n",
    "            counts = np.zeros(shape=(gridx,gridy))\n",
    "            xs = []\n",
    "            ys = []\n",
    "            thetas = []\n",
    "            nbl+=1\n",
    "            continue\n",
    "        spt = [float(x) for x in line.split()]\n",
    "        xs.append(spt[0])\n",
    "        ys.append(spt[1])\n",
    "        thetas.append(spt[2])\n",
    "\n",
    "    fin.close()\n",
    "    fout.close()\n",
    "print \"Done compiling training set\""
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [],
   "source": [
    "def plotLine(x1,y1,x2,y2):\n",
    "    plt.gca().plot([x1, x2], [y1, y2], color='b', linestyle='-', linewidth=0.2)"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAXcAAAFpCAYAAABnHGgVAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nlBCi4q4oihJCVNwVRVFCyP8PVgL8ECd+7VMAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x2af8eb6dbf10>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "iSnap = 1\n",
    "dx,dy = 1./gridx, 1./gridy\n",
    "\n",
    "f = plt.figure()\n",
    "# a = plt.Axes(f, [0., 0., 1., 1.])\n",
    "# a.set_axis_off()\n",
    "# f.add_axes(a)\n",
    "# plt.xlim(-w/2,w/2)\n",
    "# plt.ylim(-h/2,h/2)\n",
    "# plt.xticks([])\n",
    "# plt.yticks([])\n",
    "\n",
    "\n",
    "dfile = open(trnlist[1]+file_append, \"r\")\n",
    "iSnap = 1\n",
    "cntSnap = 0\n",
    "for line in dfile.readlines():\n",
    "    if cntSnap > iSnap: break\n",
    "    if cntSnap == iSnap:\n",
    "        l = [float(x) for x in line.split()]\n",
    "        for ix in range(gridx):\n",
    "            for iy in range(gridy):\n",
    "                i = int(gridx*ix + iy)\n",
    "                th = l[i]\n",
    "                x,y = dx*ix+dx/2., dy*iy+dy/2.\n",
    "                # Note th=0 is along the y-axis\n",
    "                x1 = x - (dx/2.)*sin(th*twopi) \n",
    "                y1 = y + (dy/2.)*cos(th*twopi)\n",
    "                x2 = x + (dx/2.)*sin(th*twopi)\n",
    "                y2 = y - (dy/2.)*cos(th*twopi)\n",
    "        \n",
    "                plotLine(x1,y1,x2,y2)\n",
    "        cntSnap += 1\n",
    "    else: \n",
    "        if line == \"\\n\": cntSnap+=1\n",
    "dfile.close()\n",
    "plt.gcf().set_size_inches(6,6)"
   ]
  }
 ],
 "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": "2.7.13"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 2
}