scipy interpolation weirdness

scipy_interpolation_weirdness.ipnb

{
 "metadata": {
  "name": ""
 },
 "nbformat": 3,
 "nbformat_minor": 0,
 "worksheets": [
  {
   "cells": [
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "import scipy\n",
      "from scipy.interpolate import interp1d"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [],
     "prompt_number": 1
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "scipy.__version__"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "metadata": {},
       "output_type": "pyout",
       "prompt_number": 2,
       "text": [
        "'0.13.3'"
       ]
      }
     ],
     "prompt_number": 2
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "x = [  8.,   8.,   8.,  10.,  10.,  10.,  12.,  12.,  12.,  14.,  14.]\n",
      "y = [ 21.   ,  22.375,  22.375,  22.375,  22.375,  22.375,  22.375,\n",
      "        22.375,  22.375,  23.5  ,  25.   ]"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [],
     "prompt_number": 3
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "import numpy as np\n",
      "import scipy\n",
      "from scipy.interpolate import interp1d\n",
      "inp = np.linspace(8, 14, 30)\n",
      "f_linear = interp1d(x, y, kind=\"linear\")\n",
      "pred_linear = f_linear(inp)\n",
      "pred_x_linear = f_linear(x)\n",
      "f_slinear = interp1d(x, y, kind=\"slinear\")\n",
      "pred_slinear = f_slinear(inp)\n",
      "pred_x_slinear = f_slinear(x)"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "output_type": "stream",
       "stream": "stderr",
       "text": [
        "/usr/lib/python2.7/dist-packages/scipy/interpolate/interpolate.py:416: RuntimeWarning: divide by zero encountered in true_divide\n",
        "  slope = (y_hi - y_lo) / (x_hi - x_lo)[:, None]\n",
        "/usr/lib/python2.7/dist-packages/scipy/interpolate/interpolate.py:419: RuntimeWarning: invalid value encountered in multiply\n",
        "  y_new = slope*(x_new - x_lo)[:, None] + y_lo\n"
       ]
      }
     ],
     "prompt_number": 4
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "plt.plot(inp, pred_linear, label=\"linear\")\n",
      "plt.plot(inp, pred_slinear, label=\"slinear\")\n",
      "plt.plot(x, pred_x_slinear, \"x\", label=\"slinear on x\", markersize=20)\n",
      "plt.plot(x, pred_x_linear, \"x\", label=\"linear on x\", markersize=20)\n",
      "\n",
      "plt.plot(x, y, \"o\", label=\"input\")\n",
      "plt.legend(loc=\"best\")\n",
      "plt.xlim(7, 15)\n",
      "plt.ylim(-1, 26)"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "metadata": {},
       "output_type": "pyout",
       "prompt_number": 20,
       "text": [
        "(-1, 26)"
       ]
      },
      {
       "metadata": {},
       "output_type": "display_data",
       "png": "iVBORw0KGgoAAAANSUhEUgAAAXEAAAD7CAYAAACc26SuAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\ntKL+AAAAAElFTkSuQmCC\n",
       "text": [
        "<matplotlib.figure.Figure at 0x7fcc4b197d10>"
       ]
      }
     ],
     "prompt_number": 20
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "pred_linear"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "metadata": {},
       "output_type": "pyout",
       "prompt_number": 6,
       "text": [
        "array([         nan,  22.375     ,  22.375     ,  22.375     ,\n",
        "        22.375     ,  22.375     ,  22.375     ,  22.375     ,\n",
        "        22.375     ,  22.375     ,  22.375     ,  22.375     ,\n",
        "        22.375     ,  22.375     ,  22.375     ,  22.375     ,\n",
        "        22.375     ,  22.375     ,  22.375     ,  22.375     ,\n",
        "        22.45258621,  22.56896552,  22.68534483,  22.80172414,\n",
        "        22.91810345,  23.03448276,  23.15086207,  23.26724138,\n",
        "        23.38362069,  23.5       ])"
       ]
      }
     ],
     "prompt_number": 6
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "pred_slinear"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "metadata": {},
       "output_type": "pyout",
       "prompt_number": 7,
       "text": [
        "array([  0.        ,  22.375     ,  22.375     ,  22.375     ,\n",
        "        22.375     ,  22.375     ,  22.375     ,  22.375     ,\n",
        "        22.375     ,  22.375     ,  22.375     ,  22.375     ,\n",
        "        22.375     ,  22.375     ,  22.375     ,  22.375     ,\n",
        "        22.375     ,  22.375     ,  22.375     ,  22.375     ,\n",
        "        22.45258621,  22.56896552,  22.68534483,  22.80172414,\n",
        "        22.91810345,  23.03448276,  23.15086207,  23.26724138,\n",
        "        23.38362069,   0.        ])"
       ]
      }
     ],
     "prompt_number": 7
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "pred_x_slinear"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "metadata": {},
       "output_type": "pyout",
       "prompt_number": 8,
       "text": [
        "array([ 0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.,  0.])"
       ]
      }
     ],
     "prompt_number": 8
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [],
     "language": "python",
     "metadata": {},
     "outputs": []
    }
   ],
   "metadata": {}
  }
 ]
}