{ "cells": [ { "cell_type": "markdown", "metadata": { "ExecuteTime": { "end_time": "2021-06-11T02:57:02.864712Z", "start_time": "2021-06-11T02:57:00.434753Z" } }, "source": [ "# Survival Analysis with Python\n", "\n", "\n", "**lifelines** is a complete survival analysis library, written in pure Python. What benefits does lifelines have?\n", "\n", "- easy installation\n", "- internal plotting methods\n", "- simple and intuitive API\n", "- handles right, left and interval censored data\n", "- contains the most popular parametric, semi-parametric and non-parametric models\n", "\n", "\n", "https://lifelines.readthedocs.io/\n", "\n", "```\n", "pip install lifelines\n", "```" ] }, { "cell_type": "markdown", "metadata": {}, "source": [ "Cheibub, José Antonio, Jennifer Gandhi, and James Raymond Vreeland. 2010. “Democracy and Dictatorship Revisited.” Public Choice, vol. 143, no. 2-1, pp. 67-101." ] }, { "cell_type": "code", "execution_count": 86, "metadata": { "ExecuteTime": { "end_time": "2021-06-11T03:47:14.098025Z", "start_time": "2021-06-11T03:47:14.069417Z" } }, "outputs": [ { "data": { "text/html": [ "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
ctrynamecowcode2politycodeun_region_nameun_continent_nameeheadleaderspellregdemocracyregimestart_yeardurationobserved
0Afghanistan700700.0Southern AsiaAsiaMohammad Zahir ShahMohammad Zahir Shah.Afghanistan.1946.1952.Mona...Non-democracyMonarchy194671
1Afghanistan700700.0Southern AsiaAsiaSardar Mohammad DaoudSardar Mohammad Daoud.Afghanistan.1953.1962.Ci...Non-democracyCivilian Dict1953101
2Afghanistan700700.0Southern AsiaAsiaMohammad Zahir ShahMohammad Zahir Shah.Afghanistan.1963.1972.Mona...Non-democracyMonarchy1963101
3Afghanistan700700.0Southern AsiaAsiaSardar Mohammad DaoudSardar Mohammad Daoud.Afghanistan.1973.1977.Ci...Non-democracyCivilian Dict197350
4Afghanistan700700.0Southern AsiaAsiaNur Mohammad TarakiNur Mohammad Taraki.Afghanistan.1978.1978.Civi...Non-democracyCivilian Dict197810
\n", "
" ], "text/plain": [ " ctryname cowcode2 politycode un_region_name un_continent_name \\\n", "0 Afghanistan 700 700.0 Southern Asia Asia \n", "1 Afghanistan 700 700.0 Southern Asia Asia \n", "2 Afghanistan 700 700.0 Southern Asia Asia \n", "3 Afghanistan 700 700.0 Southern Asia Asia \n", "4 Afghanistan 700 700.0 Southern Asia Asia \n", "\n", " ehead leaderspellreg \\\n", "0 Mohammad Zahir Shah Mohammad Zahir Shah.Afghanistan.1946.1952.Mona... \n", "1 Sardar Mohammad Daoud Sardar Mohammad Daoud.Afghanistan.1953.1962.Ci... \n", "2 Mohammad Zahir Shah Mohammad Zahir Shah.Afghanistan.1963.1972.Mona... \n", "3 Sardar Mohammad Daoud Sardar Mohammad Daoud.Afghanistan.1973.1977.Ci... \n", "4 Nur Mohammad Taraki Nur Mohammad Taraki.Afghanistan.1978.1978.Civi... \n", "\n", " democracy regime start_year duration observed \n", "0 Non-democracy Monarchy 1946 7 1 \n", "1 Non-democracy Civilian Dict 1953 10 1 \n", "2 Non-democracy Monarchy 1963 10 1 \n", "3 Non-democracy Civilian Dict 1973 5 0 \n", "4 Non-democracy Civilian Dict 1978 1 0 " ] }, "execution_count": 86, "metadata": {}, "output_type": "execute_result" } ], "source": [ "from lifelines.datasets import load_dd\n", "\n", "data = load_dd()\n", "data.head()" ] }, { "cell_type": "code", "execution_count": 5, "metadata": { "ExecuteTime": { "end_time": "2021-06-11T02:57:46.174422Z", "start_time": "2021-06-11T02:57:46.168141Z" } }, "outputs": [ { "data": { "text/plain": [ "array(['Monarchy', 'Civilian Dict', 'Military Dict', 'Parliamentary Dem',\n", " 'Presidential Dem', 'Mixed Dem'], dtype=object)" ] }, "execution_count": 5, "metadata": {}, "output_type": "execute_result" } ], "source": [ "data['regime'].unique()" ] }, { "cell_type": "code", "execution_count": 15, "metadata": { "ExecuteTime": { "end_time": "2021-06-11T03:02:46.604145Z", "start_time": "2021-06-11T03:02:46.599364Z" } }, "outputs": [ { "data": { "text/plain": [ "array(['Non-democracy', 'Democracy'], dtype=object)" ] }, "execution_count": 15, "metadata": {}, "output_type": "execute_result" } ], "source": [ "data['democracy'].unique()" ] }, { "cell_type": "code", "execution_count": 58, "metadata": { "ExecuteTime": { "end_time": "2021-06-11T03:38:26.335136Z", "start_time": "2021-06-11T03:38:26.332445Z" } }, "outputs": [], "source": [ "from lifelines import KaplanMeierFitter\n", "kmf = KaplanMeierFitter()" ] }, { "cell_type": "code", "execution_count": 59, "metadata": { "ExecuteTime": { "end_time": "2021-06-11T03:38:26.903572Z", "start_time": "2021-06-11T03:38:26.882322Z" } }, "outputs": [ { "data": { "text/plain": [ "" ] }, "execution_count": 59, "metadata": {}, "output_type": "execute_result" } ], "source": [ "T = data[\"duration\"]\n", "E = data[\"observed\"]\n", "\n", "kmf.fit(T, event_observed=E)\n" ] }, { "cell_type": "code", "execution_count": 60, "metadata": { "ExecuteTime": { "end_time": "2021-06-11T03:38:31.853733Z", "start_time": "2021-06-11T03:38:31.626423Z" } }, "outputs": [ { "data": { "image/png": "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\n", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "import numpy as np\n", "import matplotlib.pyplot as plt\n", "plt.style.use('ggplot')\n", "plt.figure(figsize = (8, 8))\n", "\n", "ax = plt.subplot(111)\n", "\n", "dem = (data[\"democracy\"] == \"Democracy\")\n", "\n", "t = np.linspace(0, 50, 51)\n", "kmf.fit(T[dem], event_observed=E[dem], timeline=t, label=\"Democratic Regimes\")\n", "ax = kmf.plot_survival_function(ax=ax)\n", "\n", "kmf.fit(T[~dem], event_observed=E[~dem], timeline=t, label=\"Non-democratic Regimes\")\n", "ax = kmf.plot_survival_function(ax=ax)\n", "\n", "plt.title(\"Lifespans of different global regimes\");" ] }, { "cell_type": "code", "execution_count": 61, "metadata": { "ExecuteTime": { "end_time": "2021-06-11T03:38:34.096406Z", "start_time": "2021-06-11T03:38:33.210218Z" } }, "outputs": [ { "data": { "image/png": "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\n", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "regime_types = data['regime'].unique()\n", "plt.figure(figsize = (12, 8))\n", "\n", "\n", "for i, regime_type in enumerate(regime_types):\n", " ax = plt.subplot(2, 3, i + 1)\n", "\n", " ix = data['regime'] == regime_type\n", " kmf.fit(T[ix], E[ix], label=regime_type)\n", " kmf.plot_survival_function(ax=ax, legend=False)\n", "\n", " plt.title(regime_type)\n", " plt.xlim(0, 50)\n", "\n", " if i==0:\n", " plt.ylabel('Frac. in power after $n$ years')\n", "\n", "plt.tight_layout()" ] }, { "cell_type": "code", "execution_count": 62, "metadata": { "ExecuteTime": { "end_time": "2021-06-11T03:38:36.206898Z", "start_time": "2021-06-11T03:38:36.202000Z" } }, "outputs": [ { "data": { "text/plain": [ "array(['Asia', 'Europe', 'Africa', 'Americas', 'Oceania'], dtype=object)" ] }, "execution_count": 62, "metadata": {}, "output_type": "execute_result" } ], "source": [ "data['un_continent_name'].unique()" ] }, { "cell_type": "code", "execution_count": 87, "metadata": { "ExecuteTime": { "end_time": "2021-06-11T03:47:22.142793Z", "start_time": "2021-06-11T03:47:22.133078Z" } }, "outputs": [ { "data": { "text/html": [ "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
Civilian DictMilitary DictMixed DemMonarchyParliamentary DemPresidential Dem
0000100
1100000
2000100
3100000
4100000
\n", "
" ], "text/plain": [ " Civilian Dict Military Dict Mixed Dem Monarchy Parliamentary Dem \\\n", "0 0 0 0 1 0 \n", "1 1 0 0 0 0 \n", "2 0 0 0 1 0 \n", "3 1 0 0 0 0 \n", "4 1 0 0 0 0 \n", "\n", " Presidential Dem \n", "0 0 \n", "1 0 \n", "2 0 \n", "3 0 \n", "4 0 " ] }, "execution_count": 87, "metadata": {}, "output_type": "execute_result" } ], "source": [ "import pandas as pd\n", "df = pd.get_dummies(data['regime'])\n", "df.head()" ] }, { "cell_type": "code", "execution_count": 88, "metadata": { "ExecuteTime": { "end_time": "2021-06-11T03:47:22.942729Z", "start_time": "2021-06-11T03:47:22.917729Z" } }, "outputs": [ { "data": { "text/html": [ "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
ctrynamecowcode2politycodeun_region_nameun_continent_nameeheadleaderspellregdemocracyregimestart_yeardurationobservedCivilian DictMilitary DictMixed DemMonarchyParliamentary DemPresidential Dem
0Afghanistan700700.0Southern AsiaAsiaMohammad Zahir ShahMohammad Zahir Shah.Afghanistan.1946.1952.Mona...Non-democracyMonarchy194671000100
1Afghanistan700700.0Southern AsiaAsiaSardar Mohammad DaoudSardar Mohammad Daoud.Afghanistan.1953.1962.Ci...Non-democracyCivilian Dict1953101100000
2Afghanistan700700.0Southern AsiaAsiaMohammad Zahir ShahMohammad Zahir Shah.Afghanistan.1963.1972.Mona...Non-democracyMonarchy1963101000100
3Afghanistan700700.0Southern AsiaAsiaSardar Mohammad DaoudSardar Mohammad Daoud.Afghanistan.1973.1977.Ci...Non-democracyCivilian Dict197350100000
4Afghanistan700700.0Southern AsiaAsiaNur Mohammad TarakiNur Mohammad Taraki.Afghanistan.1978.1978.Civi...Non-democracyCivilian Dict197810100000
.........................................................
1803Zambia551551.0Eastern AfricaAfricaLevy Patrick MwanawasaLevy Patrick Mwanawasa.Zambia.2002.2007.Civili...Non-democracyCivilian Dict200261100000
1804Zambia551551.0Eastern AfricaAfricaRupiah Bwezani BandaRupiah Bwezani Banda.Zambia.2008.2008.Civilian...Non-democracyCivilian Dict200810100000
1805Zimbabwe552552.0Eastern AfricaAfricaIan SmithIan Smith.Zimbabwe.1965.1978.Civilian DictNon-democracyCivilian Dict1965141100000
1806Zimbabwe552552.0Eastern AfricaAfricaAbel MuzorewaAbel Muzorewa.Zimbabwe.1979.1979.Civilian DictNon-democracyCivilian Dict197911100000
1807Zimbabwe552552.0Eastern AfricaAfricaRobert MugabeRobert Mugabe.Zimbabwe.1980.2008.Civilian DictNon-democracyCivilian Dict1980290100000
\n", "

1808 rows × 18 columns

\n", "
" ], "text/plain": [ " ctryname cowcode2 politycode un_region_name un_continent_name \\\n", "0 Afghanistan 700 700.0 Southern Asia Asia \n", "1 Afghanistan 700 700.0 Southern Asia Asia \n", "2 Afghanistan 700 700.0 Southern Asia Asia \n", "3 Afghanistan 700 700.0 Southern Asia Asia \n", "4 Afghanistan 700 700.0 Southern Asia Asia \n", "... ... ... ... ... ... \n", "1803 Zambia 551 551.0 Eastern Africa Africa \n", "1804 Zambia 551 551.0 Eastern Africa Africa \n", "1805 Zimbabwe 552 552.0 Eastern Africa Africa \n", "1806 Zimbabwe 552 552.0 Eastern Africa Africa \n", "1807 Zimbabwe 552 552.0 Eastern Africa Africa \n", "\n", " ehead \\\n", "0 Mohammad Zahir Shah \n", "1 Sardar Mohammad Daoud \n", "2 Mohammad Zahir Shah \n", "3 Sardar Mohammad Daoud \n", "4 Nur Mohammad Taraki \n", "... ... \n", "1803 Levy Patrick Mwanawasa \n", "1804 Rupiah Bwezani Banda \n", "1805 Ian Smith \n", "1806 Abel Muzorewa \n", "1807 Robert Mugabe \n", "\n", " leaderspellreg democracy \\\n", "0 Mohammad Zahir Shah.Afghanistan.1946.1952.Mona... Non-democracy \n", "1 Sardar Mohammad Daoud.Afghanistan.1953.1962.Ci... Non-democracy \n", "2 Mohammad Zahir Shah.Afghanistan.1963.1972.Mona... Non-democracy \n", "3 Sardar Mohammad Daoud.Afghanistan.1973.1977.Ci... Non-democracy \n", "4 Nur Mohammad Taraki.Afghanistan.1978.1978.Civi... Non-democracy \n", "... ... ... \n", "1803 Levy Patrick Mwanawasa.Zambia.2002.2007.Civili... Non-democracy \n", "1804 Rupiah Bwezani Banda.Zambia.2008.2008.Civilian... Non-democracy \n", "1805 Ian Smith.Zimbabwe.1965.1978.Civilian Dict Non-democracy \n", "1806 Abel Muzorewa.Zimbabwe.1979.1979.Civilian Dict Non-democracy \n", "1807 Robert Mugabe.Zimbabwe.1980.2008.Civilian Dict Non-democracy \n", "\n", " regime start_year duration observed Civilian Dict \\\n", "0 Monarchy 1946 7 1 0 \n", "1 Civilian Dict 1953 10 1 1 \n", "2 Monarchy 1963 10 1 0 \n", "3 Civilian Dict 1973 5 0 1 \n", "4 Civilian Dict 1978 1 0 1 \n", "... ... ... ... ... ... \n", "1803 Civilian Dict 2002 6 1 1 \n", "1804 Civilian Dict 2008 1 0 1 \n", "1805 Civilian Dict 1965 14 1 1 \n", "1806 Civilian Dict 1979 1 1 1 \n", "1807 Civilian Dict 1980 29 0 1 \n", "\n", " Military Dict Mixed Dem Monarchy Parliamentary Dem Presidential Dem \n", "0 0 0 1 0 0 \n", "1 0 0 0 0 0 \n", "2 0 0 1 0 0 \n", "3 0 0 0 0 0 \n", "4 0 0 0 0 0 \n", "... ... ... ... ... ... \n", "1803 0 0 0 0 0 \n", "1804 0 0 0 0 0 \n", "1805 0 0 0 0 0 \n", "1806 0 0 0 0 0 \n", "1807 0 0 0 0 0 \n", "\n", "[1808 rows x 18 columns]" ] }, "execution_count": 88, "metadata": {}, "output_type": "execute_result" } ], "source": [ "data = pd.concat([data, df], axis=1)\n", "data" ] }, { "cell_type": "code", "execution_count": 89, "metadata": { "ExecuteTime": { "end_time": "2021-06-11T03:47:26.398896Z", "start_time": "2021-06-11T03:47:26.394526Z" } }, "outputs": [ { "data": { "text/plain": [ "Index(['ctryname', 'cowcode2', 'politycode', 'un_region_name',\n", " 'un_continent_name', 'ehead', 'leaderspellreg', 'democracy', 'regime',\n", " 'start_year', 'duration', 'observed', 'Civilian Dict', 'Military Dict',\n", " 'Mixed Dem', 'Monarchy', 'Parliamentary Dem', 'Presidential Dem'],\n", " dtype='object')" ] }, "execution_count": 89, "metadata": {}, "output_type": "execute_result" } ], "source": [ "data.columns" ] }, { "cell_type": "code", "execution_count": 90, "metadata": { "ExecuteTime": { "end_time": "2021-06-11T03:47:28.697008Z", "start_time": "2021-06-11T03:47:28.692461Z" } }, "outputs": [], "source": [ "data['Democracy'] = [1 if i == 'Democracy' else 0 for i in data['democracy']]" ] }, { "cell_type": "code", "execution_count": 91, "metadata": { "ExecuteTime": { "end_time": "2021-06-11T03:48:19.350312Z", "start_time": "2021-06-11T03:48:19.212059Z" } }, "outputs": [ { "data": { "text/html": [ "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
modellifelines.CoxPHFitter
duration col'duration'
event col'observed'
baseline estimationbreslow
number of observations1808
number of events observed1468
partial log-likelihood-9613.93
time fit was run2021-06-11 03:48:19 UTC
\n", "
\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
coefexp(coef)se(coef)coef lower 95%coef upper 95%exp(coef) lower 95%exp(coef) upper 95%zp-log2(p)
start_year-0.001.000.00-0.000.001.001.00-0.820.411.28
Democracy0.972.650.060.851.102.343.0015.27<0.005172.57

\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
Concordance0.60
Partial AIC19231.86
log-likelihood ratio test264.70 on 2 df
-log2(p) of ll-ratio test190.94
\n", "
" ], "text/latex": [ "\\begin{tabular}{lrrrrrrrrrr}\n", "\\toprule\n", "{} & coef & exp(coef) & se(coef) & coef lower 95\\% & coef upper 95\\% & exp(coef) lower 95\\% & exp(coef) upper 95\\% & z & p & -log2(p) \\\\\n", "covariate & & & & & & & & & & \\\\\n", "\\midrule\n", "start\\_year & -0.00 & 1.00 & 0.00 & -0.00 & 0.00 & 1.00 & 1.00 & -0.82 & 0.41 & 1.28 \\\\\n", "Democracy & 0.97 & 2.65 & 0.06 & 0.85 & 1.10 & 2.34 & 3.00 & 15.27 & 0.00 & 172.57 \\\\\n", "\\bottomrule\n", "\\end{tabular}\n" ], "text/plain": [ "\n", " duration col = 'duration'\n", " event col = 'observed'\n", " baseline estimation = breslow\n", " number of observations = 1808\n", "number of events observed = 1468\n", " partial log-likelihood = -9613.93\n", " time fit was run = 2021-06-11 03:48:19 UTC\n", "\n", "---\n", " coef exp(coef) se(coef) coef lower 95% coef upper 95% exp(coef) lower 95% exp(coef) upper 95%\n", "covariate \n", "start_year -0.00 1.00 0.00 -0.00 0.00 1.00 1.00\n", "Democracy 0.97 2.65 0.06 0.85 1.10 2.34 3.00\n", "\n", " z p -log2(p)\n", "covariate \n", "start_year -0.82 0.41 1.28\n", "Democracy 15.27 <0.005 172.57\n", "---\n", "Concordance = 0.60\n", "Partial AIC = 19231.86\n", "log-likelihood ratio test = 264.70 on 2 df\n", "-log2(p) of ll-ratio test = 190.94" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "from lifelines import CoxPHFitter\n", "\n", "cph0 = CoxPHFitter()\n", "\n", "dat = data[['duration', 'observed', 'start_year','Democracy']]\n", "\n", "cph0.fit(dat, duration_col='duration', event_col='observed')\n", "\n", "cph0.print_summary() " ] }, { "cell_type": "code", "execution_count": 82, "metadata": { "ExecuteTime": { "end_time": "2021-06-11T03:44:58.381549Z", "start_time": "2021-06-11T03:44:58.210440Z" } }, "outputs": [ { "data": { "text/html": [ "
\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
modellifelines.CoxPHFitter
duration col'duration'
event col'observed'
baseline estimationbreslow
number of observations1808
number of events observed1468
partial log-likelihood-9576.63
time fit was run2021-06-11 03:44:58 UTC
\n", "
\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
coefexp(coef)se(coef)coef lower 95%coef upper 95%exp(coef) lower 95%exp(coef) upper 95%zp-log2(p)
start_year-0.001.000.00-0.010.000.991.00-1.810.073.83
Civilian Dict1.123.050.220.681.561.974.744.97<0.00520.52
Military Dict1.293.630.230.841.732.325.665.67<0.00526.05
Mixed Dem2.4311.380.231.982.887.2617.8510.60<0.00584.73
Parliamentary Dem1.987.260.221.552.424.6911.238.91<0.00560.73
Presidential Dem2.067.840.231.612.505.0212.249.06<0.00562.68

\n", "\n", "\n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", " \n", "
Concordance0.63
Partial AIC19165.27
log-likelihood ratio test339.29 on 6 df
-log2(p) of ll-ratio test230.92
\n", "
" ], "text/latex": [ "\\begin{tabular}{lrrrrrrrrrr}\n", "\\toprule\n", "{} & coef & exp(coef) & se(coef) & coef lower 95\\% & coef upper 95\\% & exp(coef) lower 95\\% & exp(coef) upper 95\\% & z & p & -log2(p) \\\\\n", "covariate & & & & & & & & & & \\\\\n", "\\midrule\n", "start\\_year & -0.00 & 1.00 & 0.00 & -0.01 & 0.00 & 0.99 & 1.00 & -1.81 & 0.07 & 3.83 \\\\\n", "Civilian Dict & 1.12 & 3.05 & 0.22 & 0.68 & 1.56 & 1.97 & 4.74 & 4.97 & 0.00 & 20.52 \\\\\n", "Military Dict & 1.29 & 3.63 & 0.23 & 0.84 & 1.73 & 2.32 & 5.66 & 5.67 & 0.00 & 26.05 \\\\\n", "Mixed Dem & 2.43 & 11.38 & 0.23 & 1.98 & 2.88 & 7.26 & 17.85 & 10.60 & 0.00 & 84.73 \\\\\n", "Parliamentary Dem & 1.98 & 7.26 & 0.22 & 1.55 & 2.42 & 4.69 & 11.23 & 8.91 & 0.00 & 60.73 \\\\\n", "Presidential Dem & 2.06 & 7.84 & 0.23 & 1.61 & 2.50 & 5.02 & 12.24 & 9.06 & 0.00 & 62.68 \\\\\n", "\\bottomrule\n", "\\end{tabular}\n" ], "text/plain": [ "\n", " duration col = 'duration'\n", " event col = 'observed'\n", " baseline estimation = breslow\n", " number of observations = 1808\n", "number of events observed = 1468\n", " partial log-likelihood = -9576.63\n", " time fit was run = 2021-06-11 03:44:58 UTC\n", "\n", "---\n", " coef exp(coef) se(coef) coef lower 95% coef upper 95% exp(coef) lower 95% exp(coef) upper 95%\n", "covariate \n", "start_year -0.00 1.00 0.00 -0.01 0.00 0.99 1.00\n", "Civilian Dict 1.12 3.05 0.22 0.68 1.56 1.97 4.74\n", "Military Dict 1.29 3.63 0.23 0.84 1.73 2.32 5.66\n", "Mixed Dem 2.43 11.38 0.23 1.98 2.88 7.26 17.85\n", "Parliamentary Dem 1.98 7.26 0.22 1.55 2.42 4.69 11.23\n", "Presidential Dem 2.06 7.84 0.23 1.61 2.50 5.02 12.24\n", "\n", " z p -log2(p)\n", "covariate \n", "start_year -1.81 0.07 3.83\n", "Civilian Dict 4.97 <0.005 20.52\n", "Military Dict 5.67 <0.005 26.05\n", "Mixed Dem 10.60 <0.005 84.73\n", "Parliamentary Dem 8.91 <0.005 60.73\n", "Presidential Dem 9.06 <0.005 62.68\n", "---\n", "Concordance = 0.63\n", "Partial AIC = 19165.27\n", "log-likelihood ratio test = 339.29 on 6 df\n", "-log2(p) of ll-ratio test = 230.92" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "dat = data[['duration', 'observed', 'start_year',\n", " 'Civilian Dict', 'Military Dict', #'Monarchy',\n", " 'Mixed Dem', 'Parliamentary Dem', 'Presidential Dem'\n", "]]\n", "\n", "from lifelines import CoxPHFitter\n", "\n", "cph = CoxPHFitter()\n", "cph.fit(dat, duration_col='duration', event_col='observed')\n", "\n", "cph.print_summary() " ] }, { "cell_type": "code", "execution_count": 84, "metadata": { "ExecuteTime": { "end_time": "2021-06-11T03:46:06.718596Z", "start_time": "2021-06-11T03:46:06.566579Z" } }, "outputs": [ { "data": { "image/png": "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\n", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "plt.figure(figsize = (12, 8))\n", "cph.plot()\n", "plt.show()\n" ] }, { "cell_type": "code", "execution_count": 85, "metadata": { "ExecuteTime": { "end_time": "2021-06-11T03:46:51.210158Z", "start_time": "2021-06-11T03:46:51.045959Z" } }, "outputs": [ { "data": { "image/png": "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\n", "text/plain": [ "
" ] }, "metadata": {}, "output_type": "display_data" } ], "source": [ "cph.plot_partial_effects_on_outcome(covariates='Presidential Dem', values=[0, 1], cmap='coolwarm');" ] }, { "cell_type": "code", "execution_count": null, "metadata": {}, "outputs": [], "source": [] } ], "metadata": { "kernelspec": { "display_name": "Python 3", "language": "python", "name": "python3" }, "language_info": { "codemirror_mode": { "name": "ipython", "version": 3 }, "file_extension": ".py", "mimetype": "text/x-python", "name": "python", "nbconvert_exporter": "python", "pygments_lexer": "ipython3", "version": "3.7.6" }, "toc": { "base_numbering": 1, "nav_menu": {}, "number_sections": false, "sideBar": true, "skip_h1_title": false, "title_cell": "Table of Contents", "title_sidebar": "Contents", "toc_cell": false, "toc_position": {}, "toc_section_display": true, "toc_window_display": false } }, "nbformat": 4, "nbformat_minor": 4 }