NT_X-corr.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {},
   "outputs": [],
   "source": [
    "import gps_particle_data\n",
    "import wget\n",
    "import os\n",
    "import numpy as np \n",
    "import time\n",
    "import json\n",
    "import sys\n",
    "import math\n",
    "from datetime import datetime, timedelta, date\n",
    "from itertools import compress\n",
    "from random import randint\n",
    "import urllib2\n",
    "import aacgmv2\n",
    "import scipy\n",
    "import matplotlib\n",
    "from matplotlib import gridspec\n",
    "import numpy as np\n",
    "import matplotlib.pyplot as plt\n",
    "%matplotlib inline"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 3,
   "metadata": {},
   "outputs": [
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAsMAAAIYCAYAAACIdeGuAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nMAAAY/0fjP2uO930nmUAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f1e55304b90>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAskAAAIYCAYAAACfV3FnAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nBUEfNwslnjr4AAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f1e531dbb50>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAskAAAIYCAYAAACfV3FnAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\ngg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f1e531fd910>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAskAAAIYCAYAAACfV3FnAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nAAAwiGQAABhEMgAADP8PSMQIpSgV6sYAAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f1e5311c2d0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAsMAAAIYCAYAAACIdeGuAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nSwwDADCWGAYAYCwxDADAWGIYAICx/h9/X7/evIHYjQAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f1e52f99ed0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAsMAAAIYCAYAAACIdeGuAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nAAAASUVORK5CYII=\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f1e52db7f50>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAsMAAAIYCAYAAACIdeGuAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f1e52eef910>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAsMAAAIYCAYAAACIdeGuAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nP9IevAAAAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f1e52eb1fd0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAsMAAAIYCAYAAACIdeGuAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f1e52f197d0>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAsMAAAIYCAYAAACIdeGuAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nAABJRU5ErkJggg==\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f1e52bea910>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    },
    {
     "data": {
      "image/png": "iVBORw0KGgoAAAANSUhEUgAAAsMAAAIYCAYAAACIdeGuAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz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\nODBZElJXAAAAAElFTkSuQmCC\n",
      "text/plain": [
       "<matplotlib.figure.Figure at 0x7f1e52ed2290>"
      ]
     },
     "metadata": {},
     "output_type": "display_data"
    }
   ],
   "source": [
    "for i in range(11):\n",
    "    fname = 'ch2_'+str(i)+'.ascii'\n",
    "    fig = plt.figure(figsize=(12, 9))\n",
    "    cor = np.loadtxt(fname)\n",
    "    plt.hist(cor, bins = np.arange(min(cor), max(cor)+0.1, 0.05))\n",
    "    plt.title(fname)\n",
    "    plt.show()"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 5,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "====================================\n",
      "Loading data for satellite  61\n",
      "File : https://www.ngdc.noaa.gov/stp/space-weather/satellite-data/satellite-systems/gps/data/ns61/ns61_111009_v1.03.ascii\n",
      "{0: array([ 0.70533264]), 1: array([ 0.9784285]), 2: array([ 1.]), 3: array([ 0.94334047]), 4: array([ 0.87260326]), 5: array([ 0.82971462]), 6: array([ 0.80686745]), 7: array([ 0.77339151]), 8: array([ 0.71633874]), 9: array([ 0.67820517]), 10: array([ 0.64980228])}\n"
     ]
    }
   ],
   "source": [
    "start_date = datetime(2011,10,8,0,0,0);\n",
    "end_date = datetime(2011,10,9,0,0,0);\n",
    "ms = gps_particle_data.meta_search();\n",
    "ms.load_data(start_date,end_date);\n",
    "\n",
    "#Print cross-cor coef between channel 2 and all other channels.\n",
    "dcor = {}\n",
    "signal = 'rate_electron_measured'\n",
    "output_data = ms.get_all_data_by_satellite()\n",
    "for i in range(11):\n",
    "    a0 = np.asarray(output_data[61][signal])[:,2]\n",
    "    ar2 = np.asarray(output_data[61][signal])[:,i]\n",
    "    cor = np.correlate(a0, ar2)/(math.sqrt(np.sum(a0**2)*np.sum(ar2**2)))\n",
    "    dcor.update({i:cor})\n",
    "print dcor"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "ch2_0.ascii [ 0.70533264] 0.16\n",
      "ch2_1.ascii [ 0.9784285] 1.00\n",
      "ch2_2.ascii [ 1.] 1.00\n",
      "ch2_3.ascii [ 0.94334047] 0.43\n",
      "ch2_4.ascii [ 0.87260326] 0.53\n",
      "ch2_5.ascii [ 0.82971462] 0.54\n",
      "ch2_6.ascii [ 0.80686745] 0.68\n",
      "ch2_7.ascii [ 0.77339151] 0.69\n",
      "ch2_8.ascii [ 0.71633874] 0.55\n",
      "ch2_9.ascii [ 0.67820517] 0.47\n",
      "ch2_10.ascii [ 0.64980228] 0.55\n"
     ]
    }
   ],
   "source": [
    "#To find porbability of a give X-correlation parameter to land in a given or lower bin.\n",
    "\n",
    "xv = dcor.values()\n",
    "for i in range(len(xv)):\n",
    "    sum_bins_up_to = []\n",
    "    fname = 'ch2_'+str(i)+'.ascii'\n",
    "    cor = np.loadtxt(fname)\n",
    "    sort = np.histogram(cor, bins = np.arange(min(cor), max(cor)+0.1, 0.05))\n",
    "    for j in range(len(sort[1])):\n",
    "        if sort[1][j] <= xv[i] <= sort[1][j+1]:\n",
    "            for k in np.arange(j+1):\n",
    "                sum_bins_up_to.append(sort[0][k])\n",
    "            tot_sum_up_to = np.sum(sum_bins_up_to)\n",
    "            pvalue = float(tot_sum_up_to)/float(np.sum(sort[0]))\n",
    "            print fname, xv[i], '%.2f' % (pvalue)\n",
    "#prints file name, cross-cor value and corresponding p-value"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 9,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "====================================\n",
      "Loading data for satellite  61\n",
      "File : https://www.ngdc.noaa.gov/stp/space-weather/satellite-data/satellite-systems/gps/data/ns61/ns61_080511_v1.03.ascii\n",
      "{0: array([ 0.96636995]), 1: array([ 0.98674831]), 2: array([ 1.]), 3: array([ 0.98077971]), 4: array([ 0.90766768]), 5: array([ 0.84283521]), 6: array([ 0.80268519]), 7: array([ 0.75879152]), 8: array([ 0.74644427]), 9: array([ 0.84446094]), 10: array([ 0.80657441])}\n"
     ]
    }
   ],
   "source": [
    "start_date = datetime(2008,5,8,0,0,0);\n",
    "end_date = datetime(2008,5,11,0,0,0);\n",
    "ms = gps_particle_data.meta_search();\n",
    "ms.load_data(start_date,end_date);\n",
    "\n",
    "#Print cross-cor coef between channel 2 and all other channels.\n",
    "dcor = {}\n",
    "signal = 'rate_electron_measured'\n",
    "output_data = ms.get_all_data_by_satellite()\n",
    "for i in range(11):\n",
    "    a0 = np.asarray(output_data[61][signal])[:,2]\n",
    "    ar2 = np.asarray(output_data[61][signal])[:,i]\n",
    "    cor = np.correlate(a0, ar2)/(math.sqrt(np.sum(a0**2)*np.sum(ar2**2)))\n",
    "    dcor.update({i:cor})\n",
    "print dcor"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 10,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "ch2_0.ascii [ 0.96636995] 0.87\n",
      "ch2_1.ascii [ 0.98674831] 1.00\n",
      "ch2_2.ascii [ 1.] 1.00\n",
      "ch2_3.ascii [ 0.98077971] 1.00\n",
      "ch2_4.ascii [ 0.90766768] 0.73\n",
      "ch2_5.ascii [ 0.84283521] 0.70\n",
      "ch2_6.ascii [ 0.80268519] 0.68\n",
      "ch2_7.ascii [ 0.75879152] 0.69\n",
      "ch2_8.ascii [ 0.74644427] 0.55\n",
      "ch2_9.ascii [ 0.84446094] 0.79\n",
      "ch2_10.ascii [ 0.80657441] 0.79\n"
     ]
    }
   ],
   "source": [
    "#To find porbability of a give X-correlation parameter to land in a given or lower bin.\n",
    "\n",
    "xv = dcor.values()\n",
    "for i in range(len(xv)):\n",
    "    sum_bins_up_to = []\n",
    "    fname = 'ch2_'+str(i)+'.ascii'\n",
    "    cor = np.loadtxt(fname)\n",
    "    sort = np.histogram(cor, bins = np.arange(min(cor), max(cor)+0.1, 0.05))\n",
    "    for j in range(len(sort[1])):\n",
    "        if sort[1][j] <= xv[i] <= sort[1][j+1]:\n",
    "            for k in np.arange(j+1):\n",
    "                sum_bins_up_to.append(sort[0][k])\n",
    "            tot_sum_up_to = np.sum(sum_bins_up_to)\n",
    "            pvalue = float(tot_sum_up_to)/float(np.sum(sort[0]))\n",
    "            print fname, xv[i], '%.2f' % (pvalue)\n",
    "#prints file name, cross-cor value and corresponding p-value"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 14,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "====================================\n",
      "Loading data for satellite  61\n",
      "File : https://www.ngdc.noaa.gov/stp/space-weather/satellite-data/satellite-systems/gps/data/ns61/ns61_070513_v1.03.ascii\n",
      "{0: array([ 0.80902212]), 1: array([ 0.92366296]), 2: array([ 1.]), 3: array([ 0.93019246]), 4: array([ 0.82782372]), 5: array([ 0.77896469]), 6: array([ 0.7496024]), 7: array([ 0.71691587]), 8: array([ 0.76265288]), 9: array([ 0.75994717]), 10: array([ 0.71755885])}\n"
     ]
    }
   ],
   "source": [
    "start_date = datetime(2007,5,12,0,0,0);\n",
    "end_date = datetime(2007,5,16,0,0,0);\n",
    "ms = gps_particle_data.meta_search();\n",
    "ms.load_data(start_date,end_date);\n",
    "\n",
    "#Print cross-cor coef between channel 2 and all other channels.\n",
    "dcor = {}\n",
    "signal = 'rate_electron_measured'\n",
    "output_data = ms.get_all_data_by_satellite()\n",
    "for i in range(11):\n",
    "    a0 = np.asarray(output_data[61][signal])[:,2]\n",
    "    ar2 = np.asarray(output_data[61][signal])[:,i]\n",
    "    cor = np.correlate(a0, ar2)/(math.sqrt(np.sum(a0**2)*np.sum(ar2**2)))\n",
    "    dcor.update({i:cor})\n",
    "print dcor"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 15,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "ch2_0.ascii [ 0.80902212] 0.27\n",
      "ch2_1.ascii [ 0.92366296] 0.42\n",
      "ch2_2.ascii [ 1.] 1.00\n",
      "ch2_3.ascii [ 0.93019246] 0.43\n",
      "ch2_4.ascii [ 0.82782372] 0.38\n",
      "ch2_5.ascii [ 0.77896469] 0.42\n",
      "ch2_6.ascii [ 0.7496024] 0.57\n",
      "ch2_7.ascii [ 0.71691587] 0.60\n",
      "ch2_8.ascii [ 0.76265288] 0.69\n",
      "ch2_9.ascii [ 0.75994717] 0.65\n",
      "ch2_10.ascii [ 0.71755885] 0.64\n"
     ]
    }
   ],
   "source": [
    "#To find porbability of a give X-correlation parameter to land in a given or lower bin.\n",
    "\n",
    "xv = dcor.values()\n",
    "for i in range(len(xv)):\n",
    "    sum_bins_up_to = []\n",
    "    fname = 'ch2_'+str(i)+'.ascii'\n",
    "    cor = np.loadtxt(fname)\n",
    "    sort = np.histogram(cor, bins = np.arange(min(cor), max(cor)+0.1, 0.05))\n",
    "    for j in range(len(sort[1])):\n",
    "        if sort[1][j] <= xv[i] <= sort[1][j+1]:\n",
    "            for k in np.arange(j+1):\n",
    "                sum_bins_up_to.append(sort[0][k])\n",
    "            tot_sum_up_to = np.sum(sum_bins_up_to)\n",
    "            pvalue = float(tot_sum_up_to)/float(np.sum(sort[0]))\n",
    "            print fname, xv[i], '%.2f' % (pvalue)\n",
    "#prints file name, cross-cor value and corresponding p-value"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 16,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "====================================\n",
      "Loading data for satellite  61\n",
      "File : https://www.ngdc.noaa.gov/stp/space-weather/satellite-data/satellite-systems/gps/data/ns61/ns61_110515_v1.03.ascii\n",
      "{0: array([ 0.95564481]), 1: array([ 0.98427338]), 2: array([ 1.]), 3: array([ 0.93022824]), 4: array([ 0.87148698]), 5: array([ 0.83755936]), 6: array([ 0.81558211]), 7: array([ 0.79434781]), 8: array([ 0.78319761]), 9: array([ 0.74813148]), 10: array([ 0.75177514])}\n"
     ]
    }
   ],
   "source": [
    "start_date = datetime(2011,5,12,0,0,0);\n",
    "end_date = datetime(2011,5,16,0,0,0);\n",
    "ms = gps_particle_data.meta_search();\n",
    "ms.load_data(start_date,end_date);\n",
    "\n",
    "#Print cross-cor coef between channel 2 and all other channels.\n",
    "dcor = {}\n",
    "signal = 'rate_electron_measured'\n",
    "output_data = ms.get_all_data_by_satellite()\n",
    "for i in range(11):\n",
    "    a0 = np.asarray(output_data[61][signal])[:,2]\n",
    "    ar2 = np.asarray(output_data[61][signal])[:,i]\n",
    "    cor = np.correlate(a0, ar2)/(math.sqrt(np.sum(a0**2)*np.sum(ar2**2)))\n",
    "    dcor.update({i:cor})\n",
    "print dcor"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 17,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "ch2_0.ascii [ 0.95564481] 0.87\n",
      "ch2_1.ascii [ 0.98427338] 1.00\n",
      "ch2_2.ascii [ 1.] 1.00\n",
      "ch2_3.ascii [ 0.93022824] 0.43\n",
      "ch2_4.ascii [ 0.87148698] 0.53\n",
      "ch2_5.ascii [ 0.83755936] 0.70\n",
      "ch2_6.ascii [ 0.81558211] 0.68\n",
      "ch2_7.ascii [ 0.79434781] 0.69\n",
      "ch2_8.ascii [ 0.78319761] 0.69\n",
      "ch2_9.ascii [ 0.74813148] 0.54\n",
      "ch2_10.ascii [ 0.75177514] 0.72\n"
     ]
    }
   ],
   "source": [
    "#To find porbability of a give X-correlation parameter to land in a given or lower bin.\n",
    "\n",
    "xv = dcor.values()\n",
    "for i in range(len(xv)):\n",
    "    sum_bins_up_to = []\n",
    "    fname = 'ch2_'+str(i)+'.ascii'\n",
    "    cor = np.loadtxt(fname)\n",
    "    sort = np.histogram(cor, bins = np.arange(min(cor), max(cor)+0.1, 0.05))\n",
    "    for j in range(len(sort[1])):\n",
    "        if sort[1][j] <= xv[i] <= sort[1][j+1]:\n",
    "            for k in np.arange(j+1):\n",
    "                sum_bins_up_to.append(sort[0][k])\n",
    "            tot_sum_up_to = np.sum(sum_bins_up_to)\n",
    "            pvalue = float(tot_sum_up_to)/float(np.sum(sort[0]))\n",
    "            print fname, xv[i], '%.2f' % (pvalue)\n",
    "#prints file name, cross-cor value and corresponding p-value"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "====================================\n",
      "Loading data for satellite  61\n",
      "File : https://www.ngdc.noaa.gov/stp/space-weather/satellite-data/satellite-systems/gps/data/ns61/ns61_151115_v1.03.ascii\n",
      "{0: array([ 0.98276877]), 1: array([ 0.99112553]), 2: array([ 1.]), 3: array([ 0.9941383]), 4: array([ 0.98453504]), 5: array([ 0.97196994]), 6: array([ 0.95551628]), 7: array([ 0.92798976]), 8: array([ 0.91112949]), 9: array([ 0.93291446]), 10: array([ 0.9421716])}\n"
     ]
    }
   ],
   "source": [
    "start_date = datetime(2015,11,12,0,0,0);\n",
    "end_date = datetime(2015,11,16,0,0,0);\n",
    "ms = gps_particle_data.meta_search();\n",
    "ms.load_data(start_date,end_date);\n",
    "\n",
    "#Print cross-cor coef between channel 2 and all other channels.\n",
    "dcor = {}\n",
    "signal = 'rate_electron_measured'\n",
    "output_data = ms.get_all_data_by_satellite()\n",
    "for i in range(11):\n",
    "    a0 = np.asarray(output_data[61][signal])[:,2]\n",
    "    ar2 = np.asarray(output_data[61][signal])[:,i]\n",
    "    cor = np.correlate(a0, ar2)/(math.sqrt(np.sum(a0**2)*np.sum(ar2**2)))\n",
    "    dcor.update({i:cor})\n",
    "print dcor"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "ch2_0.ascii [ 0.98276877] 1.00\n",
      "ch2_1.ascii [ 0.99112553] 1.00\n",
      "ch2_2.ascii [ 1.] 1.00\n",
      "ch2_3.ascii [ 0.9941383] 1.00\n",
      "ch2_4.ascii [ 0.98453504] 1.00\n",
      "ch2_5.ascii [ 0.97196994] 0.98\n",
      "ch2_6.ascii [ 0.95551628] 1.00\n",
      "ch2_7.ascii [ 0.92798976] 0.97\n",
      "ch2_8.ascii [ 0.91112949] 1.00\n",
      "ch2_9.ascii [ 0.93291446] 0.97\n",
      "ch2_10.ascii [ 0.9421716] 0.96\n"
     ]
    }
   ],
   "source": [
    "#To find porbability of a give X-correlation parameter to land in a given or lower bin.\n",
    "\n",
    "xv = dcor.values()\n",
    "for i in range(len(xv)):\n",
    "    sum_bins_up_to = []\n",
    "    fname = 'ch2_'+str(i)+'.ascii'\n",
    "    cor = np.loadtxt(fname)\n",
    "    sort = np.histogram(cor, bins = np.arange(min(cor), max(cor)+0.1, 0.05))\n",
    "    for j in range(len(sort[1])):\n",
    "        if sort[1][j] <= xv[i] <= sort[1][j+1]:\n",
    "            for k in np.arange(j+1):\n",
    "                sum_bins_up_to.append(sort[0][k])\n",
    "            tot_sum_up_to = np.sum(sum_bins_up_to)\n",
    "            pvalue = float(tot_sum_up_to)/float(np.sum(sort[0]))\n",
    "            print fname, xv[i], '%.2f' % (pvalue)\n",
    "#prints file name, cross-cor value and corresponding p-value"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 20,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "\n",
      "====================================\n",
      "Loading data for satellite  61\n",
      "File : https://www.ngdc.noaa.gov/stp/space-weather/satellite-data/satellite-systems/gps/data/ns61/ns61_051113_v1.03.ascii\n",
      "{0: array([ 0.95610265]), 1: array([ 0.98396681]), 2: array([ 1.]), 3: array([ 0.94019048]), 4: array([ 0.82344832]), 5: array([ 0.77314078]), 6: array([ 0.73388569]), 7: array([ 0.67130305]), 8: array([ 0.65931156]), 9: array([ 0.78584265]), 10: array([ 0.77901069])}\n"
     ]
    }
   ],
   "source": [
    "start_date = datetime(2005,11,12,0,0,0);\n",
    "end_date = datetime(2005,11,16,0,0,0);\n",
    "ms = gps_particle_data.meta_search();\n",
    "ms.load_data(start_date,end_date);\n",
    "\n",
    "#Print cross-cor coef between channel 2 and all other channels.\n",
    "dcor = {}\n",
    "signal = 'rate_electron_measured'\n",
    "output_data = ms.get_all_data_by_satellite()\n",
    "for i in range(11):\n",
    "    a0 = np.asarray(output_data[61][signal])[:,2]\n",
    "    ar2 = np.asarray(output_data[61][signal])[:,i]\n",
    "    cor = np.correlate(a0, ar2)/(math.sqrt(np.sum(a0**2)*np.sum(ar2**2)))\n",
    "    dcor.update({i:cor})\n",
    "print dcor"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 21,
   "metadata": {},
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "ch2_0.ascii [ 0.95610265] 0.87\n",
      "ch2_1.ascii [ 0.98396681] 1.00\n",
      "ch2_2.ascii [ 1.] 1.00\n",
      "ch2_3.ascii [ 0.94019048] 0.43\n",
      "ch2_4.ascii [ 0.82344832] 0.38\n",
      "ch2_5.ascii [ 0.77314078] 0.42\n",
      "ch2_6.ascii [ 0.73388569] 0.45\n",
      "ch2_7.ascii [ 0.67130305] 0.49\n",
      "ch2_8.ascii [ 0.65931156] 0.47\n",
      "ch2_9.ascii [ 0.78584265] 0.65\n",
      "ch2_10.ascii [ 0.77901069] 0.72\n"
     ]
    }
   ],
   "source": [
    "#To find porbability of a give X-correlation parameter to land in a given or lower bin.\n",
    "\n",
    "xv = dcor.values()\n",
    "for i in range(len(xv)):\n",
    "    sum_bins_up_to = []\n",
    "    fname = 'ch2_'+str(i)+'.ascii'\n",
    "    cor = np.loadtxt(fname)\n",
    "    sort = np.histogram(cor, bins = np.arange(min(cor), max(cor)+0.1, 0.05))\n",
    "    for j in range(len(sort[1])):\n",
    "        if sort[1][j] <= xv[i] <= sort[1][j+1]:\n",
    "            for k in np.arange(j+1):\n",
    "                sum_bins_up_to.append(sort[0][k])\n",
    "            tot_sum_up_to = np.sum(sum_bins_up_to)\n",
    "            pvalue = float(tot_sum_up_to)/float(np.sum(sort[0]))\n",
    "            print fname, xv[i], '%.2f' % (pvalue)\n",
    "#prints file name, cross-cor value and corresponding p-value"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": null,
   "metadata": {
    "collapsed": true
   },
   "outputs": [],
   "source": []
  }
 ],
 "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
}