Skip to content

Instantly share code, notes, and snippets.

@metasyn

metasyn/gist:51bc4f49af0d32df3c6a

Last active August 29, 2015 14:19
Show Gist options
  • Save metasyn/51bc4f49af0d32df3c6a to your computer and use it in GitHub Desktop.
Save metasyn/51bc4f49af0d32df3c6a to your computer and use it in GitHub Desktop.
Download ZIP
LDI
Raw
gistfile1.txt
{
"cells": [
{
"cell_type": "markdown",
"metadata": {},
"source": [
"# [Linguistic Diversity Index](http://en.wikipedia.org/wiki/Linguistic_diversity_index)\n",
"\n",
">Greenberg's Diversity Index (LDI) is the probability that two people selected from the population at random will have different mother tongues; it therefore ranges from 0 (everyone has the same mother tongue) to 1 (no two people have the same mother tongue).\n",
"\n",
"So we got them from Wikipedia and let's take a look. \n",
"\n",
"Made by Xander Johnson (metasyn.pw)"
]
},
{
"cell_type": "code",
"execution_count": 197,
"metadata": {
"collapsed": false
},
"outputs": [],
"source": [
"import pandas as pd\n",
"import numpy as np\n",
"import seaborn as sns\n",
"import plotly.plotly as py \n",
"from plotly.graph_objs import *\n",
"\n",
"% matplotlib inline"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Scrape"
]
},
{
"cell_type": "code",
"execution_count": 4,
"metadata": {
"collapsed": false
},
"outputs": [
{
"data": {
"application/javascript": [
"\n",
"// http://en.wikipedia.org/wiki/Linguistic_diversity_index\n",
"\n",
"// initalize\n",
"var array = []\n",
"\n",
"// get those values with our good friend jQuery\n",
"$.each($('table tbody tr'), function() {\n",
" var ldi = $(this).find('td:last').html()\n",
" array.push({country: $(this).find('a').html(), \n",
" ldi: parseFloat(ldi) ? ldi:\"None\"});\n",
"});\n",
"\n",
"// simulate download\n",
"var stringed = JSON.stringify(array)\n",
"var b = new Blob([stringed])\n",
"var url = URL.createObjectURL(b)\n",
"var a = document.createElement('a');\n",
"a.href = url;\n",
"document.body.appendChild(a);\n",
"a.click();\n",
"\n",
"// copy & paste"
],
"text/plain": [
"<IPython.core.display.Javascript object>"
]
},
"metadata": {},
"output_type": "display_data"
}
],
"source": [
"%%javascript\n",
"\n",
"// http://en.wikipedia.org/wiki/Linguistic_diversity_index"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"## Clean"
]
},
{
"cell_type": "code",
"execution_count": 5,
"metadata": {
"collapsed": false
},
"outputs": [],
"source": [
"data = [{\"country\":\"Papua New Guinea\",\"ldi\":\".990\"},{\"country\":\"Vanuatu\",\"ldi\":\".972\"},{\"country\":\"Solomon Islands\",\"ldi\":\".965\"},{\"country\":\"Tanzania\",\"ldi\":\".965\"},{\"country\":\"Central African Republic\",\"ldi\":\".960\"},{\"country\":\"Chad\",\"ldi\":\".950\"},{\"country\":\"Democratic Republic of the Congo\",\"ldi\":\".948\"},{\"country\":\"Cameroon\",\"ldi\":\".942\"},{\"country\":\"India\",\"ldi\":\".930\"},{\"country\":\"Mozambique\",\"ldi\":\".929\"},{\"country\":\"Uganda\",\"ldi\":\".928\"},{\"country\":\"Gabon\",\"ldi\":\".919\"},{\"country\":\"Côte d'Ivoire\",\"ldi\":\".917\"},{\"country\":\"Liberia\",\"ldi\":\".912\"},{\"country\":\"Angola\",\"ldi\":\".901\"},{\"country\":\"Kenya\",\"ldi\":\".901\"},{\"country\":\"Togo\",\"ldi\":\".897\"},{\"country\":\"Timor-Leste\",\"ldi\":\".897\"},{\"country\":\"Mali\",\"ldi\":\".876\"},{\"country\":\"Nigeria\",\"ldi\":\".870\"},{\"country\":\"South Africa\",\"ldi\":\".869\"},{\"country\":\"Zambia\",\"ldi\":\".855\"},{\"country\":\"Guinea-Bissau\",\"ldi\":\".853\"},{\"country\":\"Philippines\",\"ldi\":\".849\"},{\"country\":\"Bhutan\",\"ldi\":\".846\"},{\"country\":\"Indonesia\",\"ldi\":\".846\"},{\"country\":\"Ethiopia\",\"ldi\":\".843\"},{\"country\":\"Congo\",\"ldi\":\".820\"},{\"country\":\"Sierra Leone\",\"ldi\":\".817\"},{\"country\":\"Namibia\",\"ldi\":\".808\"},{\"country\":\"Ghana\",\"ldi\":\".805\"},{\"country\":\"Iran\",\"ldi\":\".797\"},{\"country\":\"Federated States of Micronesia\",\"ldi\":\".792\"},{\"country\":\"Suriname\",\"ldi\":\".788\"},{\"country\":\"Benin\",\"ldi\":\".785\"},{\"country\":\"United Arab Emirates\",\"ldi\":\".777\"},{\"country\":\"Burkina Faso\",\"ldi\":\".773\"},{\"country\":\"Senegal\",\"ldi\":\".772\"},{\"country\":\"Pakistan\",\"ldi\":\".762\"},{\"country\":\"Malaysia\",\"ldi\":\".758\"},{\"country\":\"Thailand\",\"ldi\":\".753\"},{\"country\":\"Eritrea\",\"ldi\":\".749\"},{\"country\":\"Gambia\",\"ldi\":\".748\"},{\"country\":\"Guinea\",\"ldi\":\".748\"},{\"country\":\"Singapore\",\"ldi\":\".748\"},{\"country\":\"Nepal\",\"ldi\":\".742\"},{\"country\":\"Belgium\",\"ldi\":\".734\"},{\"country\":\"Afghanistan\",\"ldi\":\".732\"},{\"country\":\"Kazakhstan\",\"ldi\":\".701\"},{\"country\":\"Trinidad and Tobago\",\"ldi\":\".696\"},{\"country\":\"Belize\",\"ldi\":\".693\"},{\"country\":\"Oman\",\"ldi\":\".693\"},{\"country\":\"Guatemala\",\"ldi\":\".691\"},{\"country\":\"Bolivia\",\"ldi\":\".680\"},{\"country\":\"Laos\",\"ldi\":\".678\"},{\"country\":\"Kyrgyzstan\",\"ldi\":\".670\"},{\"country\":\"Jordan\",\"ldi\":\".666\"},{\"country\":\"Israel\",\"ldi\":\".665\"},{\"country\":\"Bahrain\",\"ldi\":\".663\"},{\"country\":\"Norway\",\"ldi\":\".657\"},{\"country\":\"Madagascar\",\"ldi\":\".656\"},{\"country\":\"Niger\",\"ldi\":\".646\"},{\"country\":\"Mauritius\",\"ldi\":\".641\"},{\"country\":\"Saudi Arabia\",\"ldi\":\".609\"},{\"country\":\"Qatar\",\"ldi\":\".608\"},{\"country\":\"Fiji\",\"ldi\":\".607\"},{\"country\":\"Nauru\",\"ldi\":\".596\"},{\"country\":\"Latvia\",\"ldi\":\".595\"},{\"country\":\"Italy\",\"ldi\":\".593\"},{\"country\":\"Djibouti\",\"ldi\":\".592\"},{\"country\":\"Moldova\",\"ldi\":\".589\"},{\"country\":\"Sudan\",\"ldi\":\".587\"},{\"country\":\"Yemen\",\"ldi\":\".579\"},{\"country\":\"Georgia\",\"ldi\":\".576\"},{\"country\":\"Andorra\",\"ldi\":\".574\"},{\"country\":\"Macedonia\",\"ldi\":\".566\"},{\"country\":\"Kuwait\",\"ldi\":\".556\"},{\"country\":\"Comoros\",\"ldi\":\".551\"},{\"country\":\"Canada\",\"ldi\":\".549\"},{\"country\":\"Switzerland\",\"ldi\":\".547\"},{\"country\":\"Cayman Islands\",\"ldi\":\".547\"},{\"country\":\"Austria\",\"ldi\":\".540\"},{\"country\":\"Zimbabwe\",\"ldi\":\".526\"},{\"country\":\"Monaco\",\"ldi\":\".521\"},{\"country\":\"Myanmar\",\"ldi\":\".521\"},{\"country\":\"Malawi\",\"ldi\":\".519\"},{\"country\":\"Egypt\",\"ldi\":\".509\"},{\"country\":\"Syrian Arab Republic\",\"ldi\":\".503\"},{\"country\":\"Luxembourg\",\"ldi\":\".498\"},{\"country\":\"Gibraltar\",\"ldi\":\".498\"},{\"country\":\"San Marino\",\"ldi\":\".494\"},{\"country\":\"Ukraine\",\"ldi\":\".492\"},{\"country\":\"China\",\"ldi\":\".491\"},{\"country\":\"Iraq\",\"ldi\":\".484\"},{\"country\":\"Tajikistan\",\"ldi\":\".482\"},{\"country\":\"Estonia\",\"ldi\":\".476\"},{\"country\":\"Morocco\",\"ldi\":\".466\"},{\"country\":\"Brunei Darussalam\",\"ldi\":\".456\"},{\"country\":\"Equatorial Guinea\",\"ldi\":\".453\"},{\"country\":\"Botswana\",\"ldi\":\".444\"},{\"country\":\"Spain\",\"ldi\":\".438\"},{\"country\":\"Uzbekistan\",\"ldi\":\".428\"},{\"country\":\"Bosnia and Herzegovina\",\"ldi\":\".416\"},{\"country\":\"Belarus\",\"ldi\":\".397\"},{\"country\":\"Netherlands\",\"ldi\":\".389\"},{\"country\":\"Sao Tome and Principe\",\"ldi\":\".389\"},{\"country\":\"Aruba\",\"ldi\":\".387\"},{\"country\":\"Bahamas\",\"ldi\":\".386\"},{\"country\":\"Turkmenistan\",\"ldi\":\".386\"},{\"country\":\"Cook Islands\",\"ldi\":\".379\"},{\"country\":\"Peru\",\"ldi\":\".376\"},{\"country\":\"Azerbaijan\",\"ldi\":\".373\"},{\"country\":\"Cyprus\",\"ldi\":\".366\"},{\"country\":\"Libyan Arab Jamahiriya\",\"ldi\":\".362\"},{\"country\":\"Serbia\",\"ldi\":\".359\"},{\"country\":\"United States of America\",\"ldi\":\".353\"},{\"country\":\"Paraguay\",\"ldi\":\".347\"},{\"country\":\"Lithuania\",\"ldi\":\".339\"},{\"country\":\"Bangladesh\",\"ldi\":\".332\"},{\"country\":\"Mongolia\",\"ldi\":\".331\"},{\"country\":\"Panama\",\"ldi\":\".324\"},{\"country\":\"Dominica\",\"ldi\":\".313\"},{\"country\":\"Algeria\",\"ldi\":\".313\"},{\"country\":\"Sri Lanka\",\"ldi\":\".313\"},{\"country\":\"Slovakia\",\"ldi\":\".307\"},{\"country\":\"Turkey\",\"ldi\":\".289\"},{\"country\":\"Russian Federation\",\"ldi\":\".283\"},{\"country\":\"France\",\"ldi\":\".272\"},{\"country\":\"Netherlands Antilles\",\"ldi\":\".266\"},{\"country\":\"Ecuador\",\"ldi\":\".264\"},{\"country\":\"Lesotho\",\"ldi\":\".260\"},{\"country\":\"Albania\",\"ldi\":\".257\"},{\"country\":\"Viet Nam\",\"ldi\":\".234\"},{\"country\":\"Swaziland\",\"ldi\":\".228\"},{\"country\":\"Bulgaria\",\"ldi\":\".224\"},{\"country\":\"Ireland\",\"ldi\":\".223\"},{\"country\":\"Argentina\",\"ldi\":\".213\"},{\"country\":\"Palestine\",\"ldi\":\".208\"},{\"country\":\"Germany\",\"ldi\":\".189\"},{\"country\":\"Somalia\",\"ldi\":\".179\"},{\"country\":\"Greece\",\"ldi\":\".175\"},{\"country\":\"Slovenia\",\"ldi\":\".174\"},{\"country\":\"Armenia\",\"ldi\":\".174\"},{\"country\":\"Mauritania\",\"ldi\":\".172\"},{\"country\":\"Romania\",\"ldi\":\".168\"},{\"country\":\"Sweden\",\"ldi\":\".167\"},{\"country\":\"British Virgin Islands\",\"ldi\":\".167\"},{\"country\":\"Lebanon\",\"ldi\":\".161\"},{\"country\":\"Hungary\",\"ldi\":\".158\"},{\"country\":\"Cambodia\",\"ldi\":\".157\"},{\"country\":\"Turks and Caicos Islands\",\"ldi\":\".145\"},{\"country\":\"Anguilla\",\"ldi\":\".140\"},{\"country\":\"Finland\",\"ldi\":\".140\"},{\"country\":\"Tuvalu\",\"ldi\":\".139\"},{\"country\":\"United Kingdom\",\"ldi\":\".139\"},{\"country\":\"Mexico\",\"ldi\":\".135\"},{\"country\":\"Liechtenstein\",\"ldi\":\".128\"},{\"country\":\"Australia\",\"ldi\":\".126\"},{\"country\":\"New Zealand\",\"ldi\":\".102\"},{\"country\":\"Uruguay\",\"ldi\":\".092\"},{\"country\":\"Barbados\",\"ldi\":\".091\"},{\"country\":\"Croatia\",\"ldi\":\".087\"},{\"country\":\"Nicaragua\",\"ldi\":\".081\"},{\"country\":\"Guyana\",\"ldi\":\".078\"},{\"country\":\"Palau\",\"ldi\":\".077\"},{\"country\":\"Niue\",\"ldi\":\".071\"},{\"country\":\"Cape Verde\",\"ldi\":\".070\"},{\"country\":\"Czech Republic\",\"ldi\":\".069\"},{\"country\":\"Seychelles\",\"ldi\":\".067\"},{\"country\":\"Grenada\",\"ldi\":\".064\"},{\"country\":\"Poland\",\"ldi\":\".060\"},{\"country\":\"Antigua and Barbuda\",\"ldi\":\".057\"},{\"country\":\"Honduras\",\"ldi\":\".056\"},{\"country\":\"Tokelau\",\"ldi\":\".054\"},{\"country\":\"Dominican Republic\",\"ldi\":\".053\"},{\"country\":\"Denmark\",\"ldi\":\".051\"},{\"country\":\"Costa Rica\",\"ldi\":\".050\"},{\"country\":\"Chile\",\"ldi\":\".034\"},{\"country\":\"Kiribati\",\"ldi\":\".033\"},{\"country\":\"Brazil\",\"ldi\":\".032\"},{\"country\":\"Colombia\",\"ldi\":\".030\"},{\"country\":\"Japan\",\"ldi\":\".028\"},{\"country\":\"Marshall Islands\",\"ldi\":\".027\"},{\"country\":\"Montserrat\",\"ldi\":\".026\"},{\"country\":\"Venezuela\",\"ldi\":\".026\"},{\"country\":\"Portugal\",\"ldi\":\".022\"},{\"country\":\"Saint Lucia\",\"ldi\":\".020\"},{\"country\":\"Iceland\",\"ldi\":\".019\"},{\"country\":\"Malta\",\"ldi\":\".016\"},{\"country\":\"Tonga\",\"ldi\":\".014\"},{\"country\":\"Tunisia\",\"ldi\":\".012\"},{\"country\":\"Jamaica\",\"ldi\":\".011\"},{\"country\":\"Saint Kitts and Nevis\",\"ldi\":\".010\"},{\"country\":\"Maldives\",\"ldi\":\".010\"},{\"country\":\"Saint Vincent and the Grenadines\",\"ldi\":\".009\"},{\"country\":\"El Salvador\",\"ldi\":\".004\"},{\"country\":\"Burundi\",\"ldi\":\".004\"},{\"country\":\"Rwanda\",\"ldi\":\".004\"},{\"country\":\"South Korea\",\"ldi\":\".003\"},{\"country\":\"Samoa\",\"ldi\":\".002\"},{\"country\":\"Cuba\",\"ldi\":\".001\"},{\"country\":\"Haiti\",\"ldi\":\"None\"},{\"country\":\"Saint Helena\",\"ldi\":\"None\"},{\"country\":\"Holy See\",\"ldi\":\"None\"},{\"country\":\"Montenegro\",\"ldi\":\"None\"},{\"country\":\"Bermuda\",\"ldi\":\"None\"},{\"country\":\"Hong Kong\",\"ldi\":\"None\"},{\"country\":\"Macao\",\"ldi\":\"None\"},{\"country\":\"North Korea\",\"ldi\":\"None\"}]\n",
"for d in data:\n",
" d['country'] = d['country'].decode('utf-8')\n",
" if \"Ivoire\" in d['country']:\n",
" d['country'] = u\"Cote d'Ivoire\"\n",
" if d['ldi'] == 'None':\n",
" d['ldi'] = 0\n",
" else:\n",
" d['ldi'] = float(d['ldi'])\n",
" # damn cote"
]
},
{
"cell_type": "code",
"execution_count": 6,
"metadata": {
"collapsed": false
},
"outputs": [
{
"data": {
"text/plain": [
"{'country': u'Solomon Islands', 'ldi': 0.965}"
]
},
"execution_count": 6,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"data[2]"
]
},
{
"cell_type": "code",
"execution_count": 7,
"metadata": {
"collapsed": false
},
"outputs": [],
"source": [
"df = pd.DataFrame(data)"
]
},
{
"cell_type": "code",
"execution_count": 8,
"metadata": {
"collapsed": false
},
"outputs": [
{
"data": {
"text/html": [
"<div style=\"max-height:1000px;max-width:1500px;overflow:auto;\">\n",
"<table border=\"1\" class=\"dataframe\">\n",
" <thead>\n",
" <tr style=\"text-align: right;\">\n",
" <th></th>\n",
" <th>country</th>\n",
" <th>ldi</th>\n",
" </tr>\n",
" </thead>\n",
" <tbody>\n",
" <tr>\n",
" <th>0</th>\n",
" <td> Papua New Guinea</td>\n",
" <td> 0.990</td>\n",
" </tr>\n",
" <tr>\n",
" <th>1</th>\n",
" <td> Vanuatu</td>\n",
" <td> 0.972</td>\n",
" </tr>\n",
" </tbody>\n",
"</table>\n",
"</div>"
],
"text/plain": [
" country ldi\n",
"0 Papua New Guinea 0.990\n",
"1 Vanuatu 0.972"
]
},
"execution_count": 8,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"df.head(2)"
]
},
{
"cell_type": "code",
"execution_count": 9,
"metadata": {
"collapsed": false
},
"outputs": [
{
"data": {
"text/plain": [
"array([[<matplotlib.axes.AxesSubplot object at 0x10e637550>]], dtype=object)"
]
},
"execution_count": 9,
"metadata": {},
"output_type": "execute_result"
},
{
"data": {
"image/png": [
"iVBORw0KGgoAAAANSUhEUgAAAXQAAAEICAYAAABPgw/pAAAABHNCSVQICAgIfAhkiAAAAAlwSFlz\n",
"AAALEgAACxIB0t1+/AAAEuVJREFUeJzt3X9wlPWdwPHPsxtCIOSHnEV7lyOYLaCOcypmrQcG5Aw2\n",
"OrH8SBRiJBThQEYrbZg2AcwPGaqpHE49FQ8zLZQUgVERT+aEStBudYmR8kMJiBIj1qr8agzJEpJN\n",
"9rk/ZAtN9Umy7u4TP3m/Zp6ZZNfwfPgm8571+2QfDNM0BQDw7eewewAAQHgQdABQgqADgBIEHQCU\n",
"IOgAoARBBwAlCDr6hbfeeuum22+//d2uj8+fP3/riy++OEtEZMqUKXtbWloSoz8dEB4xdg8A2Mkw\n",
"DNMwDFNEZMuWLdfaPQ/wTRB09CvHjx//blFR0boTJ05899JLL/1zY2PjxcHnLr/88kBNTc3FycnJ\n",
"f7VzRiBUBB39yrJly5669tprvQ888EDZJ598ctkPf/jD/XbPBIQLe+joFwzDME3TNLxeb+a0adPW\n",
"ioikpKQ0jBs37lWbRwPChqCjX3E4HIFAIOC44PNOO+cBwomgo18wTdMwDMPMyMjYtmnTpnkiIp9/\n",
"/nlKTU3Nf9g9GxAu7KGjPzFLS0vvW7x48drbbrvt4KWXXvrJ5Zdf/rc99OBvuwDfVga3zwUAHbp9\n",
"hT516tQ9CQkJTSIiKSkpH86fP/+R4uLitQ6HIzBy5MgDZWVl9/HKBgDsZxn0tra2OBGRdevWTQw+\n",
"tmDBgv8tLCxc4na7PeXl5U9XV1dPzszM3BLpQQEA1iwvir733ntXt7a2Dp4zZ872WbNmVe/bt++G\n",
"urq6MW632yMikpGR8YrX682MzqgAACuWr9AHDRrkmzt37orc3Nxff/TRRyPnzp277cLn4+PjW5qb\n",
"m5Os/ozX/rir7Q81dbHhGDZU/rM+WbbkfnE6nXaOAQC9YfT2CyyDPmLEiPdTU1OPnPv4g+Tk5FOH\n",
"Dh362/0ufD5fQmJi4hdWf8Y7dUdi//T5d3o7V1j5T5041tHRkep0OttsHUTElBC+SUqxFuexFuex\n",
"Ft+A5ZbL5s2bZ1dUVKwUETl27Ng/+3y+hHHjxv2+trZ2goiIx+O5NT093RONQQEA1ixfoefm5v56\n",
"8eLFa/Lz8z0iIo888sjs5OTkUyUlJZV+vz/W5XIdzMrKej46owIArFgGPSYmpmPFihUzuz5eVVV1\n",
"U8QmAgCEhLf+A4ASBB0AlCDoAKAEQQcAJQg6AChB0AFACYIOAEoQdABQgqADgBIEHQCUIOgAoARB\n",
"BwAlCDoAKEHQAUAJgg4AShB0AFCCoAOAEgQdAJQg6ACgBEEHACUIOgAoQdABQAmCDgBKEHQAUIKg\n",
"A4ASBB0AlCDoAKAEQQcAJQg6AChB0AFACYIOAEoQdABQgqADgBIEHQCUIOgAoARBBwAlCDoAKNGj\n",
"oJ86dWrYhAkT/tzQ0DDq6NGj38vLy3sjPz/fU15evso0TSPSQwIAutdt0P1+/4DS0tLVgwYN8pmm\n",
"aVRUVDxWWFi4ZP369eNFxKiurp4chTkBAN3oNuiPPvroiry8vKeHDRv2mYhIXV3dGLfb7RERycjI\n",
"eMXr9WZGesgwuEREzoqIafMhfWCGvnKwFqwFa9H9WvSKZdA3b978o6FDh5648cYbf3/uIePCLZb4\n",
"+PiW5ubmpFBOHGXHRCRORAybD+kDM/SVg7VgLViL7teiV2Ksnty8efNswzDMXbt2ZR46dOiaoqKi\n",
"3zY2Nn4n+LzP50tITEz8IpQTAwDCyzLov/vd7yYEPy4oKHitvLz83hUrVqyora2dcP311//B4/Hc\n",
"esMNN1RHfkwAQHcsg96VYRhmUVHRopKSkkq/3x/rcrkOZmVlPR+p4QAAPdfjoK9bt25i8OOqqqqb\n",
"IjINACBkvLEIAJQg6ACgBEEHACUIOgAoQdABQAmCDgBKEHQAUIKgA4ASBB0AlCDoAKAEQQcAJQg6\n",
"AChB0AFACYIOAEoQdABQgqADgBIEHQCUIOgAoARBBwAlCDoAKEHQAUAJgg4AShB0AFCCoAOAEgQd\n",
"AJQg6ACgBEEHACUIOgAoQdABQAmCDgBKEHQAUIKgA4ASBB0AlCDoAKAEQQcAJQg6ACgRY/VkZ2en\n",
"s6SkpLKhoWGUYRjmQw89dG9sbGxbcXHxWofDERg5cuSBsrKy+wzDMKM1MADgq1kG/fXXX882DCOw\n",
"YcOGG2trayc89thjD4uIFBYWLnG73Z7y8vKnq6urJ2dmZm6JzrgAgK9jueVy8803v7Rs2bL5IiJ/\n",
"+ctfRiQlJTXW1dVd53a7PSIiGRkZr3i93sxoDAoAsGb5Cl1ExOl0dhYXF6/dsWPHlMcff/yON998\n",
"c1Lwufj4+Jbm5uakyI4YFpeIyFm7hziH7anzWIvzWIvzWIsvGb39gm6DLiJSUVHxo5MnT15yxx13\n",
"1La3t8cFH/f5fAmJiYlf9PakNjgmIqki0mbzHKaE8E1SirU4j7U4j7X4Biy3XF566aWZq1evXiwi\n",
"EhcX1+pwODqvuuqq3bW1tRNERDwez63p6emeaAwKALBm+Qr9Bz/4wfPFxcVr77777j90dHQMWLp0\n",
"6cK0tLT3SkpKKv1+f6zL5TqYlZX1fLSGBQB8Pcugx8XFtf7qV7+a3vXxqqqqmyI2EQAgJLyxCACU\n",
"IOgAoARBBwAlCDoAKEHQAUAJgg4AShB0AFCCoAOAEgQdAJQg6ACgBEEHACUIOgAoQdABQAmCDgBK\n",
"EHQAUIKgA4ASBB0AlCDoAKAEQQcAJQg6AChB0AFACYIOAEoQdABQgqADgBIEHQCUIOgAoARBBwAl\n",
"CDoAKEHQAUAJgg4ASsTYPQAQbZ2dnc76+nqX3XO4XK56p9PZafcc0IOgo9+pr693zVz87OHBScNs\n",
"m+FM03GpeuSu0aNGjXrftiGgDkFHvzQ4aZgMuehf7B4DCCv20AFACYIOAEqo33IxA53iazrp/OCD\n",
"D74XGxvrt3MWl8slTqfTtvP3sYuBdo8BqKM+6L6mz6XdkXDxjx/zHLBzjnMXwWTUqFG2zdDHLgba\n",
"NgOglWXQ/X7/gCVLlvzm008/TW1vbx+4YMGC5S6X61BxcfFah8MRGDly5IGysrL7DMMwozVwKLgA\n",
"dh5rAehlGfSXX345f+jQoSdWrFgxs6mp6aLJkyfvv+KKK/YWFhYucbvdnvLy8qerq6snZ2ZmbonW\n",
"wACAr2Z5UTQrK+u5Bx54oFREJBAIOGJiYvwHDx4c43a7PSIiGRkZr3i93sxoDAoAsGYZ9MGDB/vi\n",
"4+NbWlpaEhYuXPjcT37ykwcDgcDfviY+Pr6lubk5KfJjqmLaeByOwt+vJ4JzsA5fziNi789FXzpY\n",
"i79fi17p9tcWP/vss3+dNWvWzilTpqzLzs7e4HA4AsHnfD5fQmJi4hehnLgfM2w8Rkfh79cTwTlY\n",
"hy/nEbH356IvHazF369Fr1gG/eTJk5fcc889v//Zz37282nTpq0VEbnyyiv31tbWThAR8Xg8t6an\n",
"p3tCOTEAILwsL4quXr16SUtLS9KqVatKV61aVSoisnTp0oXLly//b7/fH+tyuQ5mZWU9H51RAQBW\n",
"LIO+dOnShUuXLl3Y9fGqqqqbIjYRACAkvPUfAJQg6ACgBEEHACXU38sF6IvMQEAaGhpGXPCQLTe3\n",
"4V9N+lIfu3FdyN8Pgg7YoLX5hJQ+c3L74KT64ENRf7MT/2rSeX3sxnUhfz8IOmATbpTWt2j4frCH\n",
"DgBK8Aod6Ke+Yh/fFuzjhw9BB/qpr9jHjzr28cOLoAP9mIZ9Y5zHHjoAKEHQAUAJgg4AShB0AFCC\n",
"oAOAEgQdAJTg1xYB2OZr3twU9RuV9YU3WIUDQQdgm695c1PUb1R26pND8k8pV0T7tGFH0AHYqi+8\n",
"uelM0zFbzx8u7KEDgBIEHQCUYMsFUdXlIpgt/0qPlgtgQFcEHVHV5SJY1C9+iei5AAZ0RdARdXZf\n",
"BNNyAQzoij10AFCCV+hRcm7vWMSmfWMR9o4B7Qh6lJzbO5bBSfW27BuLsHcMaEfQo4i9YwCRxB46\n",
"AChB0AFACYIOAEoQdABQgqADgBIEHQCUIOgAoARBBwAlCDoAKNGjoO/fv//7BQUFr4mIHD169Ht5\n",
"eXlv5Ofne8rLy1eZpmlEdkQAQE90G/TKysqfP/jgg5Xt7e0DRUQqKioeKywsXLJ+/frxImJUV1dP\n",
"jviUAIBudRv01NTUI08++eS04Cvxurq6MW632yMikpGR8YrX682M9JAA0E8cFhHz3NFr3Qb9lltu\n",
"2ex0OjuCn1+4xRIfH9/S3NycFMqJAQD/YLSIGOeOXuv1RVGHwxEIfuzz+RISExO/COXEAIDw6nXQ\n",
"r7zyyr21tbUTREQ8Hs+t6enpnvCPBQDorR7fD90wDFNEpKioaFFJSUml3++PdblcB7Oysp6P3HgA\n",
"gJ7qUdBTUlI+2rhx41gRkREjRnxQVVV1U0SnAgD0Gm8sAgAlCDoAKEHQAUAJgg4AShB0AFCCoAOA\n",
"EgQdAJQg6ACgBEEHACUIOgAoQdABQAmCDgBKEHQAUIKgA4ASBB0AlCDoAKAEQQcAJQg6AChB0AFA\n",
"CYIOAEoQdABQgqADgBIEHQCUIOgAoARBBwAlCDoAKEHQAUAJgg4AShB0AFCCoAOAEgQdAJQg6ACg\n",
"BEEHACUIOgAoQdABQAmCDgBKxITyRYFAwPHQQw+tOnz48L/Fxsa2LV++fO7w4cPrwz0cAKDnQnqF\n",
"vmPHjil+vz9248aNYxctWlRcUVGxMtyDAQB6J6RX6Hv27BmXkZGxTUTk6quvfuvAgQPp4R0rvM40\n",
"Hbd7BGlt/quIGMzADMzQx2boK3OEo1MhBb2lpSUxPj7+dPBzp9PZGQgEHA6HI9D1v11470xj4TeZ\n",
"EADQIyFtuQwZMuS0z+dLCH7+dTEHAERPSEEfM2bMmx6P5zYRkX379t0wevTod8I7FgCgtwzTNHv9\n",
"RaZpGsHfchERefjhh2dfdtll74d9OgBAj4UUdABA38MbiwBACYIOAEoQdABQImxBDwQCjrKysv+Z\n",
"MWOGt6Cg4LWPP/7YdeHzO3fuvD03N7d2xowZ3ueee25uuM7bF3W3Flu3bs278847a/Ly8t4oLy9/\n",
"2jRN+99ZESHdrUVQSUnJMytXrnwk2vNFU3dr8e6777rz8/M9d9111x9/+tOfbmxvb4+1a9ZI624t\n",
"Xn311ak5OTlv5+bm1m7YsOFeu+aMlv3793+/oKDgta6P97qbpmmG5di+ffu0xYsX/8Y0Tdm3b9/3\n",
"FyxYsCX4XHt7+4BJkyZ9cPr06aT29vYBOTk5tSdPnhwWrnP3tcNqLVpbWwdlZmYeOXv2bJxpmlJY\n",
"WPhsdXX17XbPbMdaBI8NGzbMnz59unflypUP2z2vXWsRCASMyZMn7/3444/TTNOUTZs2/Wd9ff1o\n",
"u2e26+di4sSJDU1NTckXtsPumSN1PPPMMz/Pzs5+Z/r06d4LHw+lm2F7hW51O4APP/zwiuHDhx9J\n",
"SEhoGjBggP+666574+233x4frnP3NVZrMXDgwLObNm3694EDB54VEeno6IiJi4trtWvWSOvuNhF7\n",
"9uwZ+84771w/ffr01abi/1MRsV6LhoaGUcnJyafWrFlTOHPmzNdPnz6dnJaWdti+aSOru5+LAQMG\n",
"+E+fPp3c1tY2yDRNwzAMtb+Ol5qaeuTJJ5+c1vXnP5Ruhi3oX3c7gOBzCQkJTcHn4uPjm1taWpLC\n",
"de6+xmotDMMwhw4dekJEpKqq6setra3xY8eO3WHXrJFmtRbHjx//7qpVq0pLS0vv1x5zEeu1aGxs\n",
"vHjv3r1j77777ifWrFmTWVNTc3NNTc1E+6aNLKu1EBGZPXv2ypycnD9lZ2cfmDhx4stDhgw5/dV/\n",
"0rffLbfcstnpdHZ0fTyUboYt6Fa3AxgyZEjThc/5fL6ExMTExnCdu6/p7tYIgUDA8ctf/vK/du3a\n",
"dfMTTzyRY8+U0WG1Ftu3b89tbGy8eN68ef9XWVlZtHXr1ru2bNlSYN+0kWW1FhdddNGp4cOHH0lL\n",
"SzscExPTkZGRsa2v3/Tum7Bai08//XT4+vXr79+5c2fqzp07R5w6deqSbdu25do3rT1C6WbYgm51\n",
"O4C0tLT3jh49OrKpqemi9vb22N27d4+/5pprdoXr3H1Nd7dGKC0tXd3e3j7wqaeemhrcetHKai1m\n",
"zpz5xAsvvJC+bt26ifPmzavIzs5+dsqUKevsmzayrNYiJSXlwzNnzgwJXhzcvXt3xqhRow7YNWuk\n",
"Wa1FW1tbnMPh6IyNjW1zOByBoUOHHm9ubk62b1p7hNLNkO62+FUmTZr0otfrnZSXl/emyJe3A9i6\n",
"dWvemTNnhtx5552VxcXFhXPmzNlumqYjJyfn18OGDfssXOfua6zW4qqrrtr9wgsv3ON2uz2zZs3a\n",
"KSJSUFDweGZm5hZ7p46M7n4uLvxvNe+TinS/Fr/4xS/mLFq06FnTNI0xY8a8OX78+FfsnjlSuluL\n",
"qVOn/nbGjBnegQMHnk1NTT0yderUtTaPHHHBn/9v0k3e+g8ASvDGIgBQgqADgBIEHQCUIOgAoARB\n",
"BwAlCDoAKPH/nGmSIf1GtikAAAAASUVORK5CYII=\n"
],
"text/plain": [
"<matplotlib.figure.Figure at 0x10e588c10>"
]
},
"metadata": {},
"output_type": "display_data"
}
],
"source": [
"df.hist()"
]
},
{
"cell_type": "code",
"execution_count": 10,
"metadata": {
"collapsed": false
},
"outputs": [
{
"data": {
"text/html": [
"<div style=\"max-height:1000px;max-width:1500px;overflow:auto;\">\n",
"<table border=\"1\" class=\"dataframe\">\n",
" <thead>\n",
" <tr style=\"text-align: right;\">\n",
" <th></th>\n",
" <th>ldi</th>\n",
" </tr>\n",
" </thead>\n",
" <tbody>\n",
" <tr>\n",
" <th>count</th>\n",
" <td> 209.000000</td>\n",
" </tr>\n",
" <tr>\n",
" <th>mean</th>\n",
" <td> 0.420153</td>\n",
" </tr>\n",
" <tr>\n",
" <th>std</th>\n",
" <td> 0.311662</td>\n",
" </tr>\n",
" <tr>\n",
" <th>min</th>\n",
" <td> 0.000000</td>\n",
" </tr>\n",
" <tr>\n",
" <th>25%</th>\n",
" <td> 0.128000</td>\n",
" </tr>\n",
" <tr>\n",
" <th>50%</th>\n",
" <td> 0.389000</td>\n",
" </tr>\n",
" <tr>\n",
" <th>75%</th>\n",
" <td> 0.691000</td>\n",
" </tr>\n",
" <tr>\n",
" <th>max</th>\n",
" <td> 0.990000</td>\n",
" </tr>\n",
" </tbody>\n",
"</table>\n",
"</div>"
],
"text/plain": [
" ldi\n",
"count 209.000000\n",
"mean 0.420153\n",
"std 0.311662\n",
"min 0.000000\n",
"25% 0.128000\n",
"50% 0.389000\n",
"75% 0.691000\n",
"max 0.990000"
]
},
"execution_count": 10,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"df.describe()"
]
},
{
"cell_type": "markdown",
"metadata": {},
"source": [
"### Click and drag to zoom, shift-click to pan, double click to autoscale."
]
},
{
"cell_type": "code",
"execution_count": 11,
"metadata": {
"collapsed": false,
"scrolled": true
},
"outputs": [
{
"name": "stderr",
"output_type": "stream",
"text": [
"/Library/Python/2.7/site-packages/requests/packages/urllib3/util/ssl_.py:79: InsecurePlatformWarning:\n",
"\n",
"A true SSLContext object is not available. This prevents urllib3 from configuring SSL appropriately and may cause certain SSL connections to fail. For more information, see https://urllib3.readthedocs.org/en/latest/security.html#insecureplatformwarning.\n",
"\n"
]
},
{
"data": {
"text/html": [
"<iframe id=\"igraph\" scrolling=\"no\" style=\"border:none;\"seamless=\"seamless\" src=\"https://plot.ly/~metasyn/27.embed\" height=\"525\" width=\"100%\"></iframe>"
],
"text/plain": [
"<plotly.tools.PlotlyDisplay object>"
]
},
"execution_count": 11,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"py.iplot([Bar(x=df.country, y=df.ldi)])"
]
},
{
"cell_type": "markdown",
"metadata": {
"collapsed": true
},
"source": [
"# LDI & GDP"
]
},
{
"cell_type": "code",
"execution_count": 256,
"metadata": {
"collapsed": false
},
"outputs": [],
"source": [
"gdp_data = \"\"\"1\t United States of America\t16,768,100\n",
"2\t China\t9,181,204\n",
"3\t Japan\t4,898,532\n",
"4\t Germany\t3,730,261\n",
"5\t France\t2,678,455\n",
"6\t United Kingdom\t2,267,456\n",
"7\t Brazil\t2,243,854\n",
"8\t Italy\t2,149,485\n",
"9\t Russia\t2,096,774\n",
"10\t India\t1,937,797\n",
"11\t Canada\t1,838,964\n",
"12\t Australia\t1,531,282\n",
"13\t Spain\t1,358,263\n",
"14\t South Korea\t1,304,554\n",
"15\t Mexico\t1,259,201\n",
"16\t Indonesia\t868,346\n",
"17\t Netherlands\t853,539\n",
"18\t Turkey\t822,149\n",
"19\t Saudi Arabia\t748,450\n",
"20\t Switzerland\t685,434\n",
"21\t Argentina\t611,726\n",
"22\t Sweden\t579,680\n",
"23\t Poland\t525,863\n",
"24\t Belgium\t524,806\n",
"25\t Norway\t522,349\n",
"26\t Nigeria\t514,965\n",
"27\t Iran\t492,783\n",
" Taiwan\t489,089\n",
"28\t Austria\t428,322\n",
"29\t Thailand\t420,167\n",
"30\t United Arab Emirates\t402,340\n",
"31\t Colombia\t378,148\n",
"32\t Venezuela\t371,339\n",
"33\t South Africa\t366,060\n",
"34\t Denmark\t336,701\n",
"35\t Malaysia\t312,434\n",
"36\t Singapore\t295,744\n",
"37\t Israel\t291,567\n",
"38\t Chile\t277,043\n",
" Hong Kong\t274,027\n",
"39\t Philippines\t272,067\n",
"40\t Finland\t267,329\n",
"41\t Egypt\t255,199\n",
"42\t Greece\t241,721\n",
"43\t Ireland\t232,077\n",
"44\t Portugal\t227,324\n",
"45\t Pakistan\t225,419\n",
"46\t Kazakhstan\t224,415\n",
"47\t Czech Republic\t208,796\n",
"48\t Algeria\t208,764\n",
"49\t Qatar\t202,450\n",
"50\t Peru\t200,269\n",
"51\t Iraq\t195,517\n",
"52\t New Zealand\t189,025\n",
"53\t Romania\t188,881\n",
"54\t Ukraine\t182,026\n",
"55\t Kuwait\t175,831\n",
"56\t Vietnam\t171,222\n",
"57\t Bangladesh\t153,505\n",
"58\t Hungary\t129,989\n",
"59\t Angola\t121,692\n",
" Puerto Rico\t105,149\n",
"60\t Morocco\t114,700\n",
"61\t Slovakia\t95,775\n",
"62\t Ecuador\t94,473\n",
"63\t Oman\t79,656\n",
"64\t Cuba\t78,694\n",
"65\t Libya\t74,597\n",
"66\t Azerbaijan\t73,557\n",
"67\t Belarus\t71,710\n",
"68\t Sri Lanka\t67,203\n",
"69\t Myanmar\t63,031\n",
"70\t Dominican Republic\t60,612\n",
"71\t Luxembourg\t60,131\n",
"72\t Croatia\t57,869\n",
"73\t Uzbekistan\t57,210\n",
"74\t Uruguay\t55,708\n",
"75\t Sudan\t54,595\n",
"76\t Bulgaria\t54,481\n",
"77\t Guatemala\t53,797\n",
" Macau\t51,753\n",
"78\t Costa Rica\t49,621\n",
"79\t Slovenia\t47,990\n",
"80\t Ghana\t47,830\n",
"81\t Lebanon\t47,221\n",
"82\t Tunisia\t46,883\n",
"83\t Lithuania\t46,403\n",
"84\t Ethiopia\t46,017\n",
"85\t Serbia\t45,520\n",
"86\t Kenya\t44,100\n",
"87\t Turkmenistan\t41,851\n",
"88\t Panama\t40,467\n",
"89\t Syria\t35,164\n",
"90\t Yemen\t34,714\n",
"91\t Jordan\t33,594\n",
"92\t Tanzania\t33,225\n",
"93\t Bahrain\t32,898\n",
"94\t Congo, Democratic Republic of the\t32,691\n",
"95\t Latvia\t30,953\n",
"96\t Bolivia\t30,601\n",
"97\t Cameroon\t29,568\n",
"98\t Paraguay\t29,208\n",
"99\t Cote d'Ivoire\t28,593\n",
"100\t Gabon\t24,076\n",
"101\t El Salvador\t23,864\n",
"102\t Trinidad and Tobago\t23,225\n",
"103\t Cyprus\t22,768\n",
"104\t Estonia\t22,376\n",
"105\t Uganda\t21,736\n",
"106\t Zambia\t21,490\n",
"107\t Afghanistan\t20,364\n",
"108\t Honduras\t18,564\n",
"109\t Nepal\t18,029\n",
"110\t Bosnia and Herzegovina\t17,319\n",
"111\t Brunei\t16,954\n",
"112\t Georgia\t15,830\n",
"113\t Papua New Guinea\t15,677\n",
"114\t Jamaica\t14,795\n",
"115\t Congo, Republic of the\t14,763\n",
"116\t Mozambique\t14,605\n",
"117\t Equatorial Guinea\t14,491\n",
"118\t North Korea\t14,411\n",
"119\t Botswana\t14,410\n",
"120\t Cambodia\t14,038\n",
"121\t Senegal\t13,962\n",
"122\t Iceland\t13,579\n",
"123\t Namibia\t12,807\n",
"124\t Albania\t12,044\n",
"125\t Mauritius\t11,452\n",
"126\t Burkina Faso\t10,687\n",
"127\t Nicaragua\t10,508\n",
"128\t Mongolia\t10,271\n",
"129\t Mali\t10,262\n",
" State of Palestine\t10,255\n",
"130\t Chad\t10,183\n",
"131\t South Sudan\t10,060\n",
"132\t Madagascar\t9,968\n",
"133\t Armenia\t9,950\n",
" New Caledonia\t9,840\n",
"134\t Zimbabwe\t9,802\n",
"135\t Macedonia, Republic of\t9,576\n",
"136\t Laos\t9,100\n",
"137\t Malta\t8,775\n",
"138\t Bahamas\t8,043\n",
"139\t Tajikistan\t7,633\n",
"140\t Benin\t7,557\n",
"141\t Moldova\t7,253\n",
"142\t Haiti\t7,187\n",
" French Polynesia\t7,150\n",
"143\t Rwanda\t7,103\n",
"144\t Niger\t6,773\n",
" Kosovo\t6,499\n",
"145\t Kyrgyzstan\t6,475\n",
"146\t Guinea\t6,092\n",
"147\t Liechtenstein\t5,827\n",
"148\t Monaco\t5,707\n",
"149\t Malawi\t5,653\n",
" Bermuda\t5,593\n",
"150\t Timor-Leste\t5,387\n",
"151\t Suriname\t5,012\n",
"152\t Barbados\t4,533\n",
"153\t Sierra Leone\t4,336\n",
"154\t Montenegro\t4,046\n",
" Netherlands Antilles\t4,009\n",
"155\t Fiji\t3,999\n",
"156\t Togo\t3,917\n",
"157\t Mauritania\t3,866\n",
"158\t Swaziland\t3,861\n",
" Cayman Islands\t3,393\n",
"159\t Andorra\t3,222\n",
"160\t Eritrea\t3,108\n",
"161\t Guyana\t2,851\n",
"162\t Maldives\t2,606\n",
" Aruba\t2,544\n",
"163\t Lesotho\t2,443\n",
" Greenland\t2,343\n",
"164\t Burundi\t2,257\n",
"165\t Central African Republic\t2,184\n",
"166\t Cape Verde\t1,903\n",
"167\t Bhutan\t1,861\n",
"168\t San Marino\t1,853\n",
"169\t Belize\t1,554\n",
"170\t Liberia\t1,491\n",
"171\t Djibouti\t1,361\n",
"172\t Saint Lucia\t1,318\n",
"173\t Somalia\t1,306\n",
"174\t Antigua and Barbuda\t1,176\n",
"175\t Seychelles\t1,031\n",
"176\t Solomon Islands\t1,010\n",
"177\t Gambia, The\t917\n",
" British Virgin Islands\t909\n",
"178\t Guinea-Bissau\t849\n",
"179\t Grenada\t783\n",
"180\t Saint Kitts and Nevis\t765\n",
" Zanzibar\t755\n",
"181\t Vanuatu\t752\n",
" Turks and Caicos Islands\t728\n",
"182\t Saint Vincent and the Grenadines\t694\n",
"183\t Samoa\t681\n",
"184\t Comoros\t616\n",
"185\t Dominica\t499\n",
"186\t Tonga\t465\n",
"187\t Micronesia, Federated States of\t327\n",
" Anguilla\t288\n",
" Cook Islands\t283\n",
"188\t Sao Tome and Principe\t261\n",
"189\t Palau\t213\n",
"190\t Marshall Islands\t198\n",
"191\t Kiribati\t176\n",
"192\t Nauru\t121\n",
" Montserrat\t64\n",
"193\t Tuvalu\t40\"\"\""
]
},
{
"cell_type": "code",
"execution_count": 257,
"metadata": {
"collapsed": true
},
"outputs": [],
"source": [
"gdp_data = gdp_data[:-1] # get rid of last new line"
]
},
{
"cell_type": "code",
"execution_count": 258,
"metadata": {
"collapsed": false
},
"outputs": [],
"source": [
"clist = [t.split('\\t')[-2].decode('utf-8').strip() for t in gdp_data.split('\\n')]"
]
},
{
"cell_type": "code",
"execution_count": 259,
"metadata": {
"collapsed": false,
"scrolled": true
},
"outputs": [],
"source": [
"gdplist = [int(t.split('\\t')[-1].replace(',','')) for t in gdp_data.split('\\n')]"
]
},
{
"cell_type": "code",
"execution_count": 260,
"metadata": {
"collapsed": false
},
"outputs": [],
"source": [
"gdp_df = pd.DataFrame()\n",
"gdp_df['gdp'] = gdplist\n",
"gdp_df['country'] = clist"
]
},
{
"cell_type": "code",
"execution_count": 261,
"metadata": {
"collapsed": false
},
"outputs": [
{
"data": {
"text/html": [
"<div style=\"max-height:1000px;max-width:1500px;overflow:auto;\">\n",
"<table border=\"1\" class=\"dataframe\">\n",
" <thead>\n",
" <tr style=\"text-align: right;\">\n",
" <th></th>\n",
" <th>gdp</th>\n",
" <th>country</th>\n",
" </tr>\n",
" </thead>\n",
" <tbody>\n",
" <tr>\n",
" <th>0</th>\n",
" <td> 16768100</td>\n",
" <td> United States of America</td>\n",
" </tr>\n",
" <tr>\n",
" <th>1</th>\n",
" <td> 9181204</td>\n",
" <td> China</td>\n",
" </tr>\n",
" <tr>\n",
" <th>2</th>\n",
" <td> 4898532</td>\n",
" <td> Japan</td>\n",
" </tr>\n",
" <tr>\n",
" <th>3</th>\n",
" <td> 3730261</td>\n",
" <td> Germany</td>\n",
" </tr>\n",
" <tr>\n",
" <th>4</th>\n",
" <td> 2678455</td>\n",
" <td> France</td>\n",
" </tr>\n",
" </tbody>\n",
"</table>\n",
"</div>"
],
"text/plain": [
" gdp country\n",
"0 16768100 United States of America\n",
"1 9181204 China\n",
"2 4898532 Japan\n",
"3 3730261 Germany\n",
"4 2678455 France"
]
},
"execution_count": 261,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"gdp_df.head()"
]
},
{
"cell_type": "code",
"execution_count": 262,
"metadata": {
"collapsed": false
},
"outputs": [],
"source": [
"mrgd = pd.merge(df, gdp_df, how='inner')"
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {
"collapsed": false
},
"outputs": [],
"source": []
},
{
"cell_type": "code",
"execution_count": 264,
"metadata": {
"collapsed": false
},
"outputs": [],
"source": [
"# normalized gdp\n",
"mrgd['n_gdp'] = (mrgd.gdp - min(mrgd.gdp)) / (max(mrgd.gdp) - min(mrgd.gdp))"
]
},
{
"cell_type": "code",
"execution_count": 265,
"metadata": {
"collapsed": false
},
"outputs": [],
"source": [
"bar_ngdp = Bar(x=mrgd.country, y=mrgd.n_gdp, name=\"GDP\")\n",
"bar_ldi = Bar(x=mrgd.country, y=mrgd.ldi, name=\"LDI\")"
]
},
{
"cell_type": "code",
"execution_count": 266,
"metadata": {
"collapsed": false
},
"outputs": [
{
"name": "stderr",
"output_type": "stream",
"text": [
"/Library/Python/2.7/site-packages/requests/packages/urllib3/util/ssl_.py:79: InsecurePlatformWarning:\n",
"\n",
"A true SSLContext object is not available. This prevents urllib3 from configuring SSL appropriately and may cause certain SSL connections to fail. For more information, see https://urllib3.readthedocs.org/en/latest/security.html#insecureplatformwarning.\n",
"\n"
]
},
{
"data": {
"text/html": [
"<iframe id=\"igraph\" scrolling=\"no\" style=\"border:none;\"seamless=\"seamless\" src=\"https://plot.ly/~metasyn/42.embed\" height=\"525\" width=\"100%\"></iframe>"
],
"text/plain": [
"<plotly.tools.PlotlyDisplay object>"
]
},
"execution_count": 266,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"py.iplot({'data': [bar_ldi, bar_ngdp], 'layout': Layout(barmode='overlay', xaxis={'tickangle': 40}, margin={'b': 150})})"
]
},
{
"cell_type": "code",
"execution_count": 278,
"metadata": {
"collapsed": false
},
"outputs": [
{
"name": "stderr",
"output_type": "stream",
"text": [
"/Library/Python/2.7/site-packages/pandas-0.15.0_93_ga0ac41d-py2.7-macosx-10.9-intel.egg/pandas/core/frame.py:1808: UserWarning:\n",
"\n",
"Boolean Series key will be reindexed to match DataFrame index.\n",
"\n"
]
}
],
"source": [
"sortd = mrgd.sort('n_gdp')[mrgd.n_gdp<.4]"
]
},
{
"cell_type": "code",
"execution_count": 279,
"metadata": {
"collapsed": true
},
"outputs": [],
"source": [
"bar_ngdp = Bar(x=sortd.country[10:], y=sortd.n_gdp[10:], name=\"GDP\")\n",
"bar_ldi = Bar(x=sortd.country[10:], y=sortd.ldi[10:], name=\"LDI\")"
]
},
{
"cell_type": "code",
"execution_count": 280,
"metadata": {
"collapsed": false
},
"outputs": [
{
"name": "stderr",
"output_type": "stream",
"text": [
"/Library/Python/2.7/site-packages/requests/packages/urllib3/util/ssl_.py:79: InsecurePlatformWarning:\n",
"\n",
"A true SSLContext object is not available. This prevents urllib3 from configuring SSL appropriately and may cause certain SSL connections to fail. For more information, see https://urllib3.readthedocs.org/en/latest/security.html#insecureplatformwarning.\n",
"\n"
]
},
{
"data": {
"text/html": [
"<iframe id=\"igraph\" scrolling=\"no\" style=\"border:none;\"seamless=\"seamless\" src=\"https://plot.ly/~metasyn/52.embed\" height=\"525\" width=\"100%\"></iframe>"
],
"text/plain": [
"<plotly.tools.PlotlyDisplay object>"
]
},
"execution_count": 280,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"py.iplot({'data': [bar_ldi, bar_ngdp], 'layout': Layout(barmode='overlay', xaxis={'tickangle': 40}, margin={'b': 150})})"
]
},
{
"cell_type": "code",
"execution_count": 281,
"metadata": {
"collapsed": false
},
"outputs": [
{
"data": {
"text/html": [
"<div style=\"max-height:1000px;max-width:1500px;overflow:auto;\">\n",
"<table border=\"1\" class=\"dataframe\">\n",
" <thead>\n",
" <tr style=\"text-align: right;\">\n",
" <th></th>\n",
" <th>ldi</th>\n",
" <th>gdp</th>\n",
" <th>n_gdp</th>\n",
" </tr>\n",
" </thead>\n",
" <tbody>\n",
" <tr>\n",
" <th>ldi</th>\n",
" <td> 1.000000</td>\n",
" <td>-0.104221</td>\n",
" <td>-0.104221</td>\n",
" </tr>\n",
" <tr>\n",
" <th>gdp</th>\n",
" <td>-0.104221</td>\n",
" <td> 1.000000</td>\n",
" <td> 1.000000</td>\n",
" </tr>\n",
" <tr>\n",
" <th>n_gdp</th>\n",
" <td>-0.104221</td>\n",
" <td> 1.000000</td>\n",
" <td> 1.000000</td>\n",
" </tr>\n",
" </tbody>\n",
"</table>\n",
"</div>"
],
"text/plain": [
" ldi gdp n_gdp\n",
"ldi 1.000000 -0.104221 -0.104221\n",
"gdp -0.104221 1.000000 1.000000\n",
"n_gdp -0.104221 1.000000 1.000000"
]
},
"execution_count": 281,
"metadata": {},
"output_type": "execute_result"
}
],
"source": [
"sortd.corr()"
]
},
{
"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": "2.7.5"
}
},
"nbformat": 4,
"nbformat_minor": 0
}
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment