Tukey-Window.ipynb

{
 "metadata": {
  "name": ""
 },
 "nbformat": 3,
 "nbformat_minor": 0,
 "worksheets": [
  {
   "cells": [
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [
      "from scipy import signal\n",
      "\n",
      "def tukeywindow(freq1, freq2, f_sample=10, width=3, sig=.1):\n",
      "    \"Erstellt einen tukey windowed Frequenz\u00fcbergang Freq1 auf Freq2 mit sig Signall\u00e4nge und entsprechende FFT\"\n",
      "    freq1, freq2, width = float(freq1), float(freq2), float(width)\n",
      "    f_sample = f_sample*1000.\n",
      "    print \"Samplingfrequenz:\", f_sample/1000, \"kHz\"\n",
      "    f1 = tile(concatenate((ones(f_sample/(2*freq1)), zeros(f_sample/(2*freq1))), axis=1), sig*freq1)\n",
      "    f2 = tile(concatenate((ones(f_sample/(2*freq2)), zeros(f_sample/(2*freq2))), axis=1), sig*freq2)\n",
      "    f = concatenate((f1, f2), axis=1)-0.5\n",
      "    window = concatenate((zeros(((width-1)/(2*width))*f.size), signal.tukey((f.size/width), sym=False)), axis=1)\n",
      "    window = concatenate((window, zeros(((width-1)/(2*width))*f.size)), axis=1)\n",
      "    window = concatenate((window, zeros(f.size-window.size)), axis=1)\n",
      "    f = f*window\n",
      "    #f = f[:502681]\n",
      "    xaxis = linspace(0, 2*sig, f.size)\n",
      "    figure()\n",
      "    plot(xaxis, f)\n",
      "    plt.xlabel('t in s')\n",
      "    plt.ylim(-1,1)\n",
      "    #plt.xlim(1,1.9)\n",
      "    figure()\n",
      "    FTdata = fft.fft(f)/f.size\n",
      "    FT = FTdata[:(FTdata.size/2)]\n",
      "    faxis = fft.fftfreq(f.size, (1./f_sample))\n",
      "    faxis = faxis[:(faxis.size/2)]\n",
      "    plot(faxis, abs(FT))\n",
      "    #plt.xlim(0, (freq1+freq2))\n",
      "    plt.xlabel('Freq in Hz')\n",
      "    plt.ylabel('Amplitude')\n",
      "    #figure()\n",
      "    #s, freqs = psd(f, f.size/3, f_sample)\n",
      "    #plt.xlim(0, (freq1+freq2))\n",
      "    return FTdata, xaxis, faxis, f_sample;\n",
      "\n",
      "\n",
      "def bandpass_ifft(freq1, freq2, f_sample, width=1, sig=6):\n",
      "    flicker, x, f, fs = tukeywindow(freq1, freq2, f_sample, width, sig)\n",
      "    lowpass = zeros((1,flicker.size))\n",
      "    lowpass[0,0:(flicker.size/fs*60)] = flicker[0:(flicker.size/fs*60)]\n",
      "    lowpass[0,(flicker.size-(flicker.size/fs*60)):] = flicker[(flicker.size-(flicker.size/fs*60)):]\n",
      "    figure()\n",
      "    plot(x, fft.ifft(lowpass*flicker.size)[0,:])\n",
      "    plt.xlim(5,7)\n",
      "    figure()\n",
      "    bpsize = (60*flicker.size/fs)+(10*flicker.size/fs)\n",
      "    bandpass = concatenate((signal.tukey(bpsize, sym=False), zeros(flicker.size-2*bpsize)), axis=1)\n",
      "    bandpass = concatenate((bandpass, signal.tukey(bpsize, sym=False)), axis=1)\n",
      "    bandpass[0:(flicker.size/fs*60)/3.] = 1.\n",
      "    bandpass[(flicker.size-(flicker.size/fs*60)/3.):] = 1.\n",
      "    bandpass = abs(flicker)*bandpass\n",
      "    sig = fft.ifft(bandpass*flicker.shape)\n",
      "    figure()\n",
      "    plot(x, real(sig))\n",
      "    plt.title('bandpass-iFFT  Flicker 200->700Hz')\n",
      "    plt.xlabel('t in s')\n",
      "    plt.xlim(5.8,6.2)\n",
      "    plt.ylim(-0.0000005, 0.0000005)\n",
      "    figure()\n",
      "    plot(f[:1000],abs(bandpass[:1000]))\n",
      "    return;\n",
      "\n",
      "bandpass_ifft(600,700,42)\n",
      "\n",
      "\n",
      "\n"
     ],
     "language": "python",
     "metadata": {},
     "outputs": [
      {
       "output_type": "stream",
       "stream": "stdout",
       "text": [
        "Samplingfrequenz: 42.0 kHz\n"
       ]
      },
      {
       "metadata": {},
       "output_type": "display_data",
       "png": "iVBORw0KGgoAAAANSUhEUgAAAX0AAAEPCAYAAACukxSbAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nAAAASUVORK5CYII=\n",
       "text": [
        "<matplotlib.figure.Figure at 0x7f65a4fc9c50>"
       ]
      },
      {
       "metadata": {},
       "output_type": "display_data",
       "png": "iVBORw0KGgoAAAANSUhEUgAAAZQAAAEPCAYAAABlZDIgAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\ngoFCRERCMFCIiEgIBgoREQnx/+K3vGMJi8FFAAAAAElFTkSuQmCC\n",
       "text": [
        "<matplotlib.figure.Figure at 0x7f65a2677a90>"
       ]
      },
      {
       "metadata": {},
       "output_type": "display_data",
       "png": "iVBORw0KGgoAAAANSUhEUgAAAYsAAAEACAYAAABCl1qQAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nYII=\n",
       "text": [
        "<matplotlib.figure.Figure at 0x7f65a2673d50>"
       ]
      },
      {
       "metadata": {},
       "output_type": "display_data",
       "text": [
        "<matplotlib.figure.Figure at 0x7f65a27eedd0>"
       ]
      },
      {
       "metadata": {},
       "output_type": "display_data",
       "png": "iVBORw0KGgoAAAANSUhEUgAAAXgAAAEZCAYAAACdNT5ZAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
       "text": [
        "<matplotlib.figure.Figure at 0x7f65a275cfd0>"
       ]
      },
      {
       "metadata": {},
       "output_type": "display_data",
       "png": "iVBORw0KGgoAAAANSUhEUgAAAZsAAAEACAYAAABlOdt4AAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nsPt/OGKRWX9L1BEAAAAASUVORK5CYII=\n",
       "text": [
        "<matplotlib.figure.Figure at 0x7f65a2449d50>"
       ]
      }
     ],
     "prompt_number": 74
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [],
     "language": "python",
     "metadata": {},
     "outputs": [],
     "prompt_number": 24
    },
    {
     "cell_type": "code",
     "collapsed": false,
     "input": [],
     "language": "python",
     "metadata": {},
     "outputs": []
    }
   ],
   "metadata": {}
  }
 ]
}