Estimation.ipynb

{
 "cells": [
  {
   "cell_type": "code",
   "execution_count": 1,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "%load_ext autoreload\n",
    "%autoreload 2\n",
    "import models\n",
    "import orca"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 2,
   "metadata": {
    "collapsed": false
   },
   "outputs": [],
   "source": [
    "%%capture\n",
    "orca.run([\"build_networks\", \"neighborhood_vars\"])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Residential Price Hedonice Estimation"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 22,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Running step 'rsh_estimate'\n",
      "REGRESSION RESULTS FOR SEGMENT 19\n",
      "\n",
      "                                 OLS Regression Results                                 \n",
      "========================================================================================\n",
      "Dep. Variable:     np.log1p(res_price_per_sqft)   R-squared:                       0.284\n",
      "Model:                                      OLS   Adj. R-squared:                  0.284\n",
      "Method:                           Least Squares   F-statistic:                     956.8\n",
      "Date:                          Fri, 04 Dec 2015   Prob (F-statistic):               0.00\n",
      "Time:                                  13:54:17   Log-Likelihood:                -50732.\n",
      "No. Observations:                         45749   AIC:                         1.015e+05\n",
      "Df Residuals:                             45729   BIC:                         1.017e+05\n",
      "Df Model:                                    19                                         \n",
      "Covariance Type:                      nonrobust                                         \n",
      "=========================================================================================================\n",
      "                                            coef    std err          t      P>|t|      [95.0% Conf. Int.]\n",
      "---------------------------------------------------------------------------------------------------------\n",
      "Intercept                               -51.8343      1.761    -29.429      0.000       -55.286   -48.382\n",
      "I(year_built < 1940)[T.True]              0.8440      0.039     21.582      0.000         0.767     0.921\n",
      "I(distance_to_coast < 10000)[T.True]      0.3130      0.018     17.348      0.000         0.278     0.348\n",
      "I(distance_to_park < 400)[T.True]         0.0364      0.015      2.362      0.018         0.006     0.067\n",
      "I(distance_to_school < 800)[T.True]       0.0997      0.012      8.072      0.000         0.076     0.124\n",
      "I(distance_to_freeway < 400)[T.True]      0.0321      0.025      1.298      0.194        -0.016     0.080\n",
      "year_built_1940to1950[T.True]             0.6356      0.031     20.405      0.000         0.575     0.697\n",
      "year_built_1950to1960[T.True]             0.5060      0.023     21.813      0.000         0.461     0.552\n",
      "year_built_1960to1970[T.True]             0.2731      0.020     13.872      0.000         0.234     0.312\n",
      "year_built_1970to1980[T.True]             0.1669      0.015     11.465      0.000         0.138     0.195\n",
      "I(ave_age_of_head_1500m < 35)[T.True]   -51.8343      1.761    -29.429      0.000       -55.286   -48.382\n",
      "ave_income                                0.0909      0.011      8.328      0.000         0.070     0.112\n",
      "jobs_400m                                -0.0988      0.003    -37.033      0.000        -0.104    -0.094\n",
      "res_occupancy_3000m                       3.0760      0.053     58.439      0.000         2.973     3.179\n",
      "np.log1p(distance_to_coast)              -0.0533      0.007     -7.563      0.000        -0.067    -0.039\n",
      "residential_units_1500m                3.008e-05   2.42e-06     12.410      0.000      2.53e-05  3.48e-05\n",
      "ave_children_1500m                       -0.4632      0.015    -30.729      0.000        -0.493    -0.434\n",
      "year_built                                0.0036      0.001      7.029      0.000         0.003     0.005\n",
      "ave_year_built_1500m                     13.0515      0.477     27.342      0.000        12.116    13.987\n",
      "np.log1p(jobs_within_30_min)              0.3117      0.006     55.881      0.000         0.301     0.323\n",
      "np.log1p(population_within_15_min)       -0.3497      0.007    -48.360      0.000        -0.364    -0.336\n",
      "==============================================================================\n",
      "Omnibus:                    25174.031   Durbin-Watson:                   0.645\n",
      "Prob(Omnibus):                  0.000   Jarque-Bera (JB):           291466.985\n",
      "Skew:                          -2.426   Prob(JB):                         0.00\n",
      "Kurtosis:                      14.374   Cond. No.                     2.28e+19\n",
      "==============================================================================\n",
      "\n",
      "Warnings:\n",
      "[1] Standard Errors assume that the covariance matrix of the errors is correctly specified.\n",
      "[2] The smallest eigenvalue is 9.31e-28. This might indicate that there are\n",
      "strong multicollinearity problems or that the design matrix is singular.\n",
      "REGRESSION RESULTS FOR SEGMENT 20\n",
      "\n",
      "                                 OLS Regression Results                                 \n",
      "========================================================================================\n",
      "Dep. Variable:     np.log1p(res_price_per_sqft)   R-squared:                       0.259\n",
      "Model:                                      OLS   Adj. R-squared:                  0.256\n",
      "Method:                           Least Squares   F-statistic:                     110.3\n",
      "Date:                          Fri, 04 Dec 2015   Prob (F-statistic):               0.00\n",
      "Time:                                  13:54:17   Log-Likelihood:                -11720.\n",
      "No. Observations:                          6980   AIC:                         2.349e+04\n",
      "Df Residuals:                              6957   BIC:                         2.364e+04\n",
      "Df Model:                                    22                                         \n",
      "Covariance Type:                      nonrobust                                         \n",
      "=========================================================================================================\n",
      "                                            coef    std err          t      P>|t|      [95.0% Conf. Int.]\n",
      "---------------------------------------------------------------------------------------------------------\n",
      "Intercept                               -86.9694     30.851     -2.819      0.005      -147.446   -26.493\n",
      "I(year_built < 1940)[T.True]              1.1159      0.304      3.668      0.000         0.520     1.712\n",
      "I(distance_to_coast < 10000)[T.True]      0.1507      0.063      2.396      0.017         0.027     0.274\n",
      "I(distance_to_park < 400)[T.True]         0.2698      0.063      4.250      0.000         0.145     0.394\n",
      "I(distance_to_school < 800)[T.True]       0.0415      0.051      0.818      0.414        -0.058     0.141\n",
      "year_built_1940to1950[T.True]             1.1979      0.237      5.055      0.000         0.733     1.662\n",
      "year_built_1950to1960[T.True]             1.1044      0.193      5.719      0.000         0.726     1.483\n",
      "year_built_1960to1970[T.True]             0.8449      0.163      5.179      0.000         0.525     1.165\n",
      "year_built_1970to1980[T.True]             0.4730      0.112      4.224      0.000         0.254     0.693\n",
      "year_built_1980to1990[T.True]             0.2313      0.077      2.987      0.003         0.079     0.383\n",
      "I(ave_age_of_head_1500m < 35)[T.True]   -86.9694     30.851     -2.819      0.005      -147.446   -26.493\n",
      "I(distance_to_freeway < 800)[T.True]     -0.3577      0.066     -5.431      0.000        -0.487    -0.229\n",
      "ave_income                               -0.3013      0.084     -3.590      0.000        -0.466    -0.137\n",
      "jobs_400m                                -0.1165      0.010    -11.719      0.000        -0.136    -0.097\n",
      "res_occupancy_3000m                       3.3757      0.251     13.424      0.000         2.883     3.869\n",
      "np.log1p(distance_to_coast)              -0.3631      0.025    -14.493      0.000        -0.412    -0.314\n",
      "residential_units_1500m                9.536e-05    8.1e-06     11.771      0.000      7.95e-05     0.000\n",
      "ave_children_1500m                        0.0994      0.073      1.367      0.172        -0.043     0.242\n",
      "np.log1p(year_built)                     13.6580      7.862      1.737      0.082        -1.754    29.070\n",
      "ave_year_built_1500m                     11.0308      1.219      9.050      0.000         8.641    13.420\n",
      "np.log1p(distance_to_onramp)             -0.1570      0.023     -6.857      0.000        -0.202    -0.112\n",
      "np.log1p(distance_to_transit)            -0.2618      0.029     -8.944      0.000        -0.319    -0.204\n",
      "np.log1p(jobs_within_30_min)              0.0912      0.035      2.568      0.010         0.022     0.161\n",
      "np.log1p(population_within_15_min)       -0.2722      0.046     -5.914      0.000        -0.362    -0.182\n",
      "==============================================================================\n",
      "Omnibus:                      433.892   Durbin-Watson:                   0.724\n",
      "Prob(Omnibus):                  0.000   Jarque-Bera (JB):              519.188\n",
      "Skew:                          -0.668   Prob(JB):                    1.82e-113\n",
      "Kurtosis:                       2.947   Cond. No.                     3.21e+19\n",
      "==============================================================================\n",
      "\n",
      "Warnings:\n",
      "[1] Standard Errors assume that the covariance matrix of the errors is correctly specified.\n",
      "[2] The smallest eigenvalue is 1.22e-28. This might indicate that there are\n",
      "strong multicollinearity problems or that the design matrix is singular.\n",
      "REGRESSION RESULTS FOR SEGMENT 21\n",
      "\n",
      "                                 OLS Regression Results                                 \n",
      "========================================================================================\n",
      "Dep. Variable:     np.log1p(res_price_per_sqft)   R-squared:                       0.267\n",
      "Model:                                      OLS   Adj. R-squared:                  0.265\n",
      "Method:                           Least Squares   F-statistic:                     137.4\n",
      "Date:                          Fri, 04 Dec 2015   Prob (F-statistic):               0.00\n",
      "Time:                                  13:54:17   Log-Likelihood:                -11423.\n",
      "No. Observations:                          6443   AIC:                         2.288e+04\n",
      "Df Residuals:                              6425   BIC:                         2.300e+04\n",
      "Df Model:                                    17                                         \n",
      "Covariance Type:                      nonrobust                                         \n",
      "=========================================================================================================\n",
      "                                            coef    std err          t      P>|t|      [95.0% Conf. Int.]\n",
      "---------------------------------------------------------------------------------------------------------\n",
      "Intercept                              -182.7499     13.940    -13.110      0.000      -210.076  -155.424\n",
      "I(year_built < 1940)[T.True]              3.1552      0.194     16.267      0.000         2.775     3.535\n",
      "I(distance_to_school < 800)[T.True]      -0.4771      0.078     -6.105      0.000        -0.630    -0.324\n",
      "year_built_1940to1950[T.True]             3.2299      0.215     14.998      0.000         2.808     3.652\n",
      "year_built_1950to1960[T.True]             2.1074      0.130     16.154      0.000         1.852     2.363\n",
      "year_built_1960to1970[T.True]             1.7127      0.087     19.668      0.000         1.542     1.883\n",
      "I(ave_age_of_head_1500m < 35)[T.True]  -182.7499     13.940    -13.110      0.000      -210.076  -155.424\n",
      "ave_income                                0.1487      0.096      1.555      0.120        -0.039     0.336\n",
      "jobs_3000m                            -3.484e-05   2.88e-06    -12.115      0.000     -4.05e-05 -2.92e-05\n",
      "res_occupancy_3000m                       0.8477      0.165      5.128      0.000         0.524     1.172\n",
      "residential_units_1500m                 7.53e-05   6.81e-06     11.063      0.000       6.2e-05  8.86e-05\n",
      "np.log1p(year_built)                     49.2831      3.674     13.412      0.000        42.080    56.486\n",
      "np.log1p(distance_to_onramp)             -0.0981      0.020     -4.793      0.000        -0.138    -0.058\n",
      "np.log1p(distance_to_transit)            -0.3464      0.029    -11.946      0.000        -0.403    -0.290\n",
      "np.log1p(distance_to_school)             -0.3252      0.047     -6.969      0.000        -0.417    -0.234\n",
      "np.log1p(distance_to_coast)              -0.2462      0.016    -15.161      0.000        -0.278    -0.214\n",
      "np.log1p(jobs_within_30_min)              0.2728      0.045      6.087      0.000         0.185     0.361\n",
      "np.log1p(population_within_15_min)       -0.2627      0.058     -4.508      0.000        -0.377    -0.148\n",
      "is_peripheral                            -0.8686      1.446     -0.601      0.548        -3.704     1.966\n",
      "==============================================================================\n",
      "Omnibus:                      353.544   Durbin-Watson:                   0.863\n",
      "Prob(Omnibus):                  0.000   Jarque-Bera (JB):              428.569\n",
      "Skew:                           0.565   Prob(JB):                     8.66e-94\n",
      "Kurtosis:                       3.566   Cond. No.                     2.42e+20\n",
      "==============================================================================\n",
      "\n",
      "Warnings:\n",
      "[1] Standard Errors assume that the covariance matrix of the errors is correctly specified.\n",
      "[2] The smallest eigenvalue is 1.7e-29. This might indicate that there are\n",
      "strong multicollinearity problems or that the design matrix is singular.\n",
      "Time to execute step 'rsh_estimate': 4.36 s\n",
      "Total time to execute iteration 1 with iteration value None: 4.36 s\n"
     ]
    }
   ],
   "source": [
    "orca.run([\"rsh_estimate\"])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 23,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Running step 'rsh_simulate'\n",
      "count    632386.000000\n",
      "mean        198.673383\n",
      "std         110.583603\n",
      "min           1.193853\n",
      "25%         133.371884\n",
      "50%         182.816569\n",
      "75%         245.252267\n",
      "max        3905.574146\n",
      "dtype: float64\n",
      "Time to execute step 'rsh_simulate': 13.58 s\n",
      "Total time to execute iteration 1 with iteration value None: 13.58 s\n"
     ]
    }
   ],
   "source": [
    "orca.run([\"rsh_simulate\"])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## Non-residential Rent Hedonic Estimation"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 18,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Running step 'nrh_estimate'\n",
      "REGRESSION RESULTS FOR SEGMENT 2\n",
      "\n",
      "                                  OLS Regression Results                                  \n",
      "==========================================================================================\n",
      "Dep. Variable:     np.log1p(nonres_rent_per_sqft)   R-squared:                       0.234\n",
      "Model:                                        OLS   Adj. R-squared:                  0.230\n",
      "Method:                             Least Squares   F-statistic:                     55.41\n",
      "Date:                            Fri, 04 Dec 2015   Prob (F-statistic):           9.30e-98\n",
      "Time:                                    13:48:45   Log-Likelihood:                -223.45\n",
      "No. Observations:                            1822   AIC:                             468.9\n",
      "Df Residuals:                                1811   BIC:                             529.5\n",
      "Df Model:                                      10                                         \n",
      "Covariance Type:                        nonrobust                                         \n",
      "======================================================================================================\n",
      "                                         coef    std err          t      P>|t|      [95.0% Conf. Int.]\n",
      "------------------------------------------------------------------------------------------------------\n",
      "Intercept                             -0.7206      1.476     -0.488      0.626        -3.616     2.175\n",
      "I(year_built < 1940)[T.True]           0.7152      0.279      2.568      0.010         0.169     1.261\n",
      "I(year_built > 2005)[T.True]           0.1352      0.031      4.427      0.000         0.075     0.195\n",
      "np.log1p(stories)                      0.3857      0.041      9.405      0.000         0.305     0.466\n",
      "jobs_400m                              0.0198      0.007      2.937      0.003         0.007     0.033\n",
      "year_built                             0.0017      0.001      2.429      0.015         0.000     0.003\n",
      "distance_to_coast                  -4.046e-06   4.79e-07     -8.452      0.000     -4.98e-06 -3.11e-06\n",
      "distance_to_freeway                -2.254e-05   2.47e-06     -9.130      0.000     -2.74e-05 -1.77e-05\n",
      "nonres_occupancy_3000m                 0.0101      0.003      4.043      0.000         0.005     0.015\n",
      "np.log1p(jobs_within_30_min)           0.1134      0.012      9.083      0.000         0.089     0.138\n",
      "np.log1p(population_within_15_min)    -0.1543      0.019     -8.101      0.000        -0.192    -0.117\n",
      "==============================================================================\n",
      "Omnibus:                      435.494   Durbin-Watson:                   1.176\n",
      "Prob(Omnibus):                  0.000   Jarque-Bera (JB):             2162.010\n",
      "Skew:                           1.039   Prob(JB):                         0.00\n",
      "Kurtosis:                       7.915   Cond. No.                     7.67e+06\n",
      "==============================================================================\n",
      "\n",
      "Warnings:\n",
      "[1] Standard Errors assume that the covariance matrix of the errors is correctly specified.\n",
      "[2] The condition number is large, 7.67e+06. This might indicate that there are\n",
      "strong multicollinearity or other numerical problems.\n",
      "REGRESSION RESULTS FOR SEGMENT 4\n",
      "\n",
      "                                  OLS Regression Results                                  \n",
      "==========================================================================================\n",
      "Dep. Variable:     np.log1p(nonres_rent_per_sqft)   R-squared:                       0.297\n",
      "Model:                                        OLS   Adj. R-squared:                  0.290\n",
      "Method:                             Least Squares   F-statistic:                     43.39\n",
      "Date:                            Fri, 04 Dec 2015   Prob (F-statistic):           5.94e-72\n",
      "Time:                                    13:48:45   Log-Likelihood:                -193.56\n",
      "No. Observations:                            1038   AIC:                             409.1\n",
      "Df Residuals:                                1027   BIC:                             463.5\n",
      "Df Model:                                      10                                         \n",
      "Covariance Type:                        nonrobust                                         \n",
      "======================================================================================================\n",
      "                                         coef    std err          t      P>|t|      [95.0% Conf. Int.]\n",
      "------------------------------------------------------------------------------------------------------\n",
      "Intercept                             -6.9374      1.819     -3.813      0.000       -10.508    -3.367\n",
      "I(year_built < 1940)[T.True]           0.1881      0.130      1.447      0.148        -0.067     0.443\n",
      "I(year_built > 2005)[T.True]           0.1784      0.047      3.788      0.000         0.086     0.271\n",
      "np.log1p(stories)                      0.1282      0.022      5.804      0.000         0.085     0.172\n",
      "ave_income                             0.0002      0.003      0.061      0.951        -0.006     0.007\n",
      "jobs_400m                              0.0111      0.008      1.365      0.172        -0.005     0.027\n",
      "year_built                             0.0059      0.001      6.625      0.000         0.004     0.008\n",
      "distance_to_coast                      -6e-06   5.81e-07    -10.328      0.000     -7.14e-06 -4.86e-06\n",
      "nonres_occupancy_3000m                -0.0206      0.005     -3.920      0.000        -0.031    -0.010\n",
      "np.log1p(jobs_within_30_min)           0.0344      0.017      1.978      0.048         0.000     0.069\n",
      "np.log1p(population_within_15_min)    -0.1719      0.018     -9.338      0.000        -0.208    -0.136\n",
      "==============================================================================\n",
      "Omnibus:                       59.111   Durbin-Watson:                   1.436\n",
      "Prob(Omnibus):                  0.000   Jarque-Bera (JB):              131.166\n",
      "Skew:                          -0.340   Prob(JB):                     3.29e-29\n",
      "Kurtosis:                       4.603   Cond. No.                     5.58e+06\n",
      "==============================================================================\n",
      "\n",
      "Warnings:\n",
      "[1] Standard Errors assume that the covariance matrix of the errors is correctly specified.\n",
      "[2] The condition number is large, 5.58e+06. This might indicate that there are\n",
      "strong multicollinearity or other numerical problems.\n",
      "REGRESSION RESULTS FOR SEGMENT 5\n",
      "\n",
      "                                  OLS Regression Results                                  \n",
      "==========================================================================================\n",
      "Dep. Variable:     np.log1p(nonres_rent_per_sqft)   R-squared:                       0.236\n",
      "Model:                                        OLS   Adj. R-squared:                  0.231\n",
      "Method:                             Least Squares   F-statistic:                     41.68\n",
      "Date:                            Fri, 04 Dec 2015   Prob (F-statistic):           3.76e-72\n",
      "Time:                                    13:48:45   Log-Likelihood:                -480.29\n",
      "No. Observations:                            1359   AIC:                             982.6\n",
      "Df Residuals:                                1348   BIC:                             1040.\n",
      "Df Model:                                      10                                         \n",
      "Covariance Type:                        nonrobust                                         \n",
      "======================================================================================================\n",
      "                                         coef    std err          t      P>|t|      [95.0% Conf. Int.]\n",
      "------------------------------------------------------------------------------------------------------\n",
      "Intercept                             -2.6492      1.433     -1.849      0.065        -5.460     0.162\n",
      "I(year_built < 1940)[T.True]           0.1786      0.068      2.619      0.009         0.045     0.312\n",
      "I(year_built > 2005)[T.True]           0.1742      0.062      2.825      0.005         0.053     0.295\n",
      "ave_income                             0.0537      0.006      8.879      0.000         0.042     0.066\n",
      "jobs_1500m                             0.0526      0.010      5.449      0.000         0.034     0.072\n",
      "year_built                             0.0030      0.001      4.286      0.000         0.002     0.004\n",
      "distance_to_coast                   -3.67e-06   6.25e-07     -5.873      0.000      -4.9e-06 -2.44e-06\n",
      "nonres_occupancy_3000m                -0.0127      0.017     -0.751      0.453        -0.046     0.020\n",
      "np.log1p(jobs_within_30_min)           0.1441      0.019      7.647      0.000         0.107     0.181\n",
      "np.log1p(population_within_15_min)    -0.2457      0.020    -12.002      0.000        -0.286    -0.206\n",
      "is_peripheral                         -0.9871      0.354     -2.790      0.005        -1.681    -0.293\n",
      "==============================================================================\n",
      "Omnibus:                       14.514   Durbin-Watson:                   1.224\n",
      "Prob(Omnibus):                  0.001   Jarque-Bera (JB):               17.046\n",
      "Skew:                          -0.174   Prob(JB):                     0.000199\n",
      "Kurtosis:                       3.425   Cond. No.                     4.95e+06\n",
      "==============================================================================\n",
      "\n",
      "Warnings:\n",
      "[1] Standard Errors assume that the covariance matrix of the errors is correctly specified.\n",
      "[2] The condition number is large, 4.95e+06. This might indicate that there are\n",
      "strong multicollinearity or other numerical problems.\n",
      "Time to execute step 'nrh_estimate': 8.21 s\n",
      "Total time to execute iteration 1 with iteration value None: 8.22 s\n"
     ]
    }
   ],
   "source": [
    "orca.run([\"nrh_estimate\"])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 6,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Running step 'nrh_simulate'\n",
      "count    27077.000000\n",
      "mean        17.510207\n",
      "std          5.402668\n",
      "min          1.013028\n",
      "25%         12.595182\n",
      "50%         17.547487\n",
      "75%         20.589631\n",
      "max         41.917346\n",
      "dtype: float64\n",
      "Time to execute step 'nrh_simulate': 10.45 s\n",
      "Total time to execute iteration 1 with iteration value None: 10.45 s\n"
     ]
    }
   ],
   "source": [
    "orca.run([\"nrh_simulate\"])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 19,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Running step 'nrh_estimate2'\n",
      "                                  OLS Regression Results                                  \n",
      "==========================================================================================\n",
      "Dep. Variable:     np.log1p(nonres_rent_per_sqft)   R-squared:                       0.224\n",
      "Model:                                        OLS   Adj. R-squared:                  0.222\n",
      "Method:                             Least Squares   F-statistic:                     101.0\n",
      "Date:                            Fri, 04 Dec 2015   Prob (F-statistic):          8.45e-221\n",
      "Time:                                    13:49:14   Log-Likelihood:                -1995.1\n",
      "No. Observations:                            4219   AIC:                             4016.\n",
      "Df Residuals:                                4206   BIC:                             4099.\n",
      "Df Model:                                      12                                         \n",
      "Covariance Type:                        nonrobust                                         \n",
      "======================================================================================================\n",
      "                                         coef    std err          t      P>|t|      [95.0% Conf. Int.]\n",
      "------------------------------------------------------------------------------------------------------\n",
      "Intercept                              2.5895      1.113      2.327      0.020         0.408     4.771\n",
      "I(year_built < 1940)[T.True]           0.1937      0.065      2.988      0.003         0.067     0.321\n",
      "I(year_built > 2005)[T.True]           0.1424      0.031      4.619      0.000         0.082     0.203\n",
      "np.log1p(stories)                      0.3168      0.021     15.338      0.000         0.276     0.357\n",
      "ave_income                             0.0187      0.002     10.327      0.000         0.015     0.022\n",
      "jobs_400m                              0.0128      0.006      2.241      0.025         0.002     0.024\n",
      "year_built                             0.0008      0.001      1.457      0.145        -0.000     0.002\n",
      "distance_to_coast                  -6.861e-06    3.8e-07    -18.050      0.000     -7.61e-06 -6.12e-06\n",
      "distance_to_freeway                -1.654e-05   2.16e-06     -7.658      0.000     -2.08e-05 -1.23e-05\n",
      "nonres_occupancy_3000m                -0.0216      0.003     -7.489      0.000        -0.027    -0.016\n",
      "np.log1p(jobs_within_30_min)           0.0451      0.011      4.079      0.000         0.023     0.067\n",
      "np.log1p(population_within_15_min)    -0.1769      0.014    -12.285      0.000        -0.205    -0.149\n",
      "is_peripheral                         -0.9252      0.394     -2.350      0.019        -1.697    -0.153\n",
      "==============================================================================\n",
      "Omnibus:                       31.500   Durbin-Watson:                   0.957\n",
      "Prob(Omnibus):                  0.000   Jarque-Bera (JB):               32.049\n",
      "Skew:                           0.213   Prob(JB):                     1.10e-07\n",
      "Kurtosis:                       3.024   Cond. No.                     5.93e+06\n",
      "==============================================================================\n",
      "\n",
      "Warnings:\n",
      "[1] Standard Errors assume that the covariance matrix of the errors is correctly specified.\n",
      "[2] The condition number is large, 5.93e+06. This might indicate that there are\n",
      "strong multicollinearity or other numerical problems.\n",
      "Time to execute step 'nrh_estimate2': 7.36 s\n",
      "Total time to execute iteration 1 with iteration value None: 7.36 s\n"
     ]
    }
   ],
   "source": [
    "orca.run([\"nrh_estimate2\"])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 8,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Running step 'nrh_simulate2'\n",
      "count    75064.000000\n",
      "mean        17.103065\n",
      "std          3.267287\n",
      "min          2.995131\n",
      "25%         15.163965\n",
      "50%         16.513385\n",
      "75%         18.458550\n",
      "max         95.234036\n",
      "dtype: float64\n",
      "Time to execute step 'nrh_simulate2': 9.12 s\n",
      "Total time to execute iteration 1 with iteration value None: 9.12 s\n"
     ]
    }
   ],
   "source": [
    "orca.run([\"nrh_simulate2\"])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## ELCM Estimation"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 25,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Running step 'nrh_simulate'\n",
      "count    27077.000000\n",
      "mean        17.510207\n",
      "std          5.402668\n",
      "min          1.013028\n",
      "25%         12.595182\n",
      "50%         17.547487\n",
      "75%         20.589631\n",
      "max         41.917346\n",
      "dtype: float64\n",
      "Time to execute step 'nrh_simulate': 10.70 s\n",
      "Total time to execute iteration 1 with iteration value None: 10.70 s\n",
      "Running step 'nrh_simulate2'\n",
      "count    75064.000000\n",
      "mean        17.103065\n",
      "std          3.267287\n",
      "min          2.995131\n",
      "25%         15.163965\n",
      "50%         16.513385\n",
      "75%         18.458550\n",
      "max         95.234036\n",
      "dtype: float64\n",
      "Time to execute step 'nrh_simulate2': 9.33 s\n",
      "Total time to execute iteration 1 with iteration value None: 9.33 s\n",
      "Running step 'elcm_estimate'\n",
      "LCM RESULTS FOR SEGMENT 1\n",
      "\n",
      "Null Log-liklihood: -4605.170\n",
      "Log-liklihood at convergence: -856.898\n",
      "Log-liklihood Ratio: 0.814\n",
      "\n",
      "+------------------------------------+-------------+------------+---------+\n",
      "| Component                          | Coefficient | Std. Error | T-Score |\n",
      "+------------------------------------+-------------+------------+---------+\n",
      "| jobs_800m                          |    0.088    |   0.038    |  2.325  |\n",
      "| ave_parcel_size                    |    1.033    |   0.052    |  19.753 |\n",
      "| is_office                          |    -0.343   |   0.326    |  -1.054 |\n",
      "| is_retail                          |    -0.415   |   0.182    |  -2.274 |\n",
      "| jobs1                              |    1.573    |   0.045    |  35.101 |\n",
      "| np.log1p(jobs_within_30_min)       |    -0.070   |   0.074    |  -0.937 |\n",
      "| np.log1p(population_within_15_min) |    -0.539   |   0.072    |  -7.444 |\n",
      "| nonres_rent_per_sqft               |    0.166    |   0.011    |  14.531 |\n",
      "| is_peripheral                      |    -2.290   |   0.485    |  -4.716 |\n",
      "+------------------------------------+-------------+------------+---------+\n",
      "LCM RESULTS FOR SEGMENT 2\n",
      "\n",
      "Null Log-liklihood: -1082.215\n",
      "Log-liklihood at convergence: -28.527\n",
      "Log-liklihood Ratio: 0.974\n",
      "\n",
      "+------------------------------------+-------------+------------+---------+\n",
      "| Component                          | Coefficient | Std. Error | T-Score |\n",
      "+------------------------------------+-------------+------------+---------+\n",
      "| jobs_800m                          |    0.300    |   0.321    |  0.934  |\n",
      "| ave_parcel_size                    |    0.915    |   0.521    |  1.757  |\n",
      "| is_office                          |    0.660    |   1.110    |  0.595  |\n",
      "| is_retail                          |    0.358    |   0.956    |  0.375  |\n",
      "| jobs2                              |    3.000    |   0.431    |  6.955  |\n",
      "| np.log1p(jobs_within_30_min)       |    0.295    |   0.707    |  0.417  |\n",
      "| np.log1p(population_within_15_min) |    -0.088   |   0.672    |  -0.131 |\n",
      "| nonres_rent_per_sqft               |    0.163    |   0.075    |  2.155  |\n",
      "| is_peripheral                      |    -3.000   |  3804.758  |  -0.001 |\n",
      "+------------------------------------+-------------+------------+---------+\n",
      "LCM RESULTS FOR SEGMENT 3\n",
      "\n",
      "Null Log-liklihood: -4605.170\n",
      "Log-liklihood at convergence: -379.818\n",
      "Log-liklihood Ratio: 0.918\n",
      "\n",
      "+------------------------------------+-------------+------------+---------+\n",
      "| Component                          | Coefficient | Std. Error | T-Score |\n",
      "+------------------------------------+-------------+------------+---------+\n",
      "| jobs_800m                          |    0.044    |   0.066    |  0.664  |\n",
      "| ave_parcel_size                    |    1.163    |   0.113    |  10.315 |\n",
      "| is_office                          |    -0.430   |   0.252    |  -1.707 |\n",
      "| is_retail                          |    -0.625   |   0.203    |  -3.079 |\n",
      "| jobs3                              |    1.638    |   0.052    |  31.456 |\n",
      "| np.log1p(jobs_within_30_min)       |    -0.540   |   0.134    |  -4.023 |\n",
      "| np.log1p(population_within_15_min) |    0.392    |   0.129    |  3.045  |\n",
      "| nonres_rent_per_sqft               |    0.158    |   0.010    |  15.393 |\n",
      "| is_peripheral                      |    -2.155   |   1.285    |  -1.677 |\n",
      "+------------------------------------+-------------+------------+---------+\n",
      "LCM RESULTS FOR SEGMENT 4\n",
      "\n",
      "Null Log-liklihood: -4605.170\n",
      "Log-liklihood at convergence: -1639.901\n",
      "Log-liklihood Ratio: 0.644\n",
      "\n",
      "+------------------------------------+-------------+------------+---------+\n",
      "| Component                          | Coefficient | Std. Error | T-Score |\n",
      "+------------------------------------+-------------+------------+---------+\n",
      "| jobs_800m                          |    -0.185   |   0.046    |  -4.017 |\n",
      "| ave_parcel_size                    |    1.183    |   0.065    |  18.321 |\n",
      "| is_office                          |    0.177    |   0.157    |  1.127  |\n",
      "| is_retail                          |    -0.412   |   0.112    |  -3.670 |\n",
      "| jobs4                              |    1.473    |   0.044    |  33.203 |\n",
      "| np.log1p(jobs_within_30_min)       |    -0.437   |   0.080    |  -5.480 |\n",
      "| np.log1p(population_within_15_min) |    0.380    |   0.073    |  5.200  |\n",
      "| nonres_rent_per_sqft               |    0.133    |   0.008    |  16.639 |\n",
      "| is_peripheral                      |    -3.000   |   0.804    |  -3.730 |\n",
      "+------------------------------------+-------------+------------+---------+\n",
      "LCM RESULTS FOR SEGMENT 5\n",
      "\n",
      "Null Log-liklihood: -4605.170\n",
      "Log-liklihood at convergence: -1072.127\n",
      "Log-liklihood Ratio: 0.767\n",
      "\n",
      "+------------------------------------+-------------+------------+---------+\n",
      "| Component                          | Coefficient | Std. Error | T-Score |\n",
      "+------------------------------------+-------------+------------+---------+\n",
      "| jobs_800m                          |    0.018    |   0.071    |  0.258  |\n",
      "| ave_parcel_size                    |    0.915    |   0.088    |  10.396 |\n",
      "| is_office                          |    -0.631   |   0.204    |  -3.091 |\n",
      "| is_retail                          |    -0.917   |   0.184    |  -4.978 |\n",
      "| jobs5                              |    1.267    |   0.055    |  22.916 |\n",
      "| np.log1p(jobs_within_30_min)       |    -0.437   |   0.094    |  -4.646 |\n",
      "| np.log1p(population_within_15_min) |    0.302    |   0.083    |  3.648  |\n",
      "| nonres_rent_per_sqft               |    0.140    |   0.013    |  10.477 |\n",
      "| is_peripheral                      |    -1.316   |   1.038    |  -1.268 |\n",
      "+------------------------------------+-------------+------------+---------+\n",
      "LCM RESULTS FOR SEGMENT 6\n",
      "\n",
      "Null Log-liklihood: -4605.170\n",
      "Log-liklihood at convergence: -1470.471\n",
      "Log-liklihood Ratio: 0.681\n",
      "\n",
      "+------------------------------------+-------------+------------+---------+\n",
      "| Component                          | Coefficient | Std. Error | T-Score |\n",
      "+------------------------------------+-------------+------------+---------+\n",
      "| jobs_800m                          |    0.024    |   0.054    |  0.442  |\n",
      "| ave_parcel_size                    |    1.124    |   0.077    |  14.627 |\n",
      "| is_office                          |    -0.286   |   0.159    |  -1.795 |\n",
      "| is_retail                          |    -0.343   |   0.122    |  -2.811 |\n",
      "| jobs6                              |    1.290    |   0.049    |  26.251 |\n",
      "| np.log1p(jobs_within_30_min)       |    -0.476   |   0.074    |  -6.477 |\n",
      "| np.log1p(population_within_15_min) |    0.441    |   0.067    |  6.587  |\n",
      "| nonres_rent_per_sqft               |    0.118    |   0.008    |  14.689 |\n",
      "| is_peripheral                      |    -3.000   |   1.433    |  -2.093 |\n",
      "+------------------------------------+-------------+------------+---------+\n",
      "LCM RESULTS FOR SEGMENT 7\n",
      "\n",
      "Null Log-liklihood: -4605.170\n",
      "Log-liklihood at convergence: -1675.979\n",
      "Log-liklihood Ratio: 0.636\n",
      "\n",
      "+------------------------------------+-------------+------------+---------+\n",
      "| Component                          | Coefficient | Std. Error | T-Score |\n",
      "+------------------------------------+-------------+------------+---------+\n",
      "| jobs_800m                          |    0.055    |   0.053    |  1.032  |\n",
      "| ave_parcel_size                    |    1.322    |   0.063    |  21.049 |\n",
      "| is_office                          |    0.722    |   0.158    |  4.564  |\n",
      "| is_retail                          |    1.155    |   0.109    |  10.576 |\n",
      "| jobs7                              |    0.997    |   0.054    |  18.387 |\n",
      "| np.log1p(jobs_within_30_min)       |    -0.395   |   0.076    |  -5.217 |\n",
      "| np.log1p(population_within_15_min) |    0.485    |   0.069    |  7.032  |\n",
      "| nonres_rent_per_sqft               |    0.116    |   0.010    |  11.272 |\n",
      "| is_peripheral                      |    -0.841   |   0.651    |  -1.293 |\n",
      "+------------------------------------+-------------+------------+---------+\n",
      "LCM RESULTS FOR SEGMENT 8\n",
      "\n",
      "Null Log-liklihood: -4605.170\n",
      "Log-liklihood at convergence: -1172.035\n",
      "Log-liklihood Ratio: 0.745\n",
      "\n",
      "+------------------------------------+-------------+------------+---------+\n",
      "| Component                          | Coefficient | Std. Error | T-Score |\n",
      "+------------------------------------+-------------+------------+---------+\n",
      "| jobs_800m                          |    0.029    |   0.045    |  0.636  |\n",
      "| ave_parcel_size                    |    1.167    |   0.056    |  20.775 |\n",
      "| is_office                          |    -0.841   |   0.188    |  -4.474 |\n",
      "| is_retail                          |    -0.573   |   0.128    |  -4.495 |\n",
      "| jobs8                              |    1.463    |   0.044    |  33.125 |\n",
      "| np.log1p(jobs_within_30_min)       |    -0.534   |   0.084    |  -6.384 |\n",
      "| np.log1p(population_within_15_min) |    0.420    |   0.074    |  5.670  |\n",
      "| nonres_rent_per_sqft               |    0.124    |   0.010    |  11.940 |\n",
      "| is_peripheral                      |    -1.810   |   0.443    |  -4.088 |\n",
      "+------------------------------------+-------------+------------+---------+\n",
      "LCM RESULTS FOR SEGMENT 9\n",
      "\n",
      "Null Log-liklihood: -4605.170\n",
      "Log-liklihood at convergence: -891.499\n",
      "Log-liklihood Ratio: 0.806\n",
      "\n",
      "+------------------------------------+-------------+------------+---------+\n",
      "| Component                          | Coefficient | Std. Error | T-Score |\n",
      "+------------------------------------+-------------+------------+---------+\n",
      "| jobs_800m                          |    -0.022   |   0.058    |  -0.373 |\n",
      "| ave_parcel_size                    |    1.259    |   0.082    |  15.365 |\n",
      "| is_office                          |    0.412    |   0.171    |  2.407  |\n",
      "| is_retail                          |    -0.413   |   0.169    |  -2.444 |\n",
      "| jobs9                              |    1.351    |   0.053    |  25.256 |\n",
      "| np.log1p(jobs_within_30_min)       |    -0.078   |   0.095    |  -0.820 |\n",
      "| np.log1p(population_within_15_min) |    0.067    |   0.088    |  0.756  |\n",
      "| nonres_rent_per_sqft               |    0.128    |   0.013    |  10.149 |\n",
      "| is_peripheral                      |    -3.000   |   1.364    |  -2.200 |\n",
      "+------------------------------------+-------------+------------+---------+\n",
      "LCM RESULTS FOR SEGMENT 10\n",
      "\n",
      "Null Log-liklihood: -4605.170\n",
      "Log-liklihood at convergence: -1182.500\n",
      "Log-liklihood Ratio: 0.743\n",
      "\n",
      "+------------------------------------+-------------+------------+---------+\n",
      "| Component                          | Coefficient | Std. Error | T-Score |\n",
      "+------------------------------------+-------------+------------+---------+\n",
      "| jobs_800m                          |    -0.005   |   0.059    |  -0.080 |\n",
      "| ave_parcel_size                    |    1.044    |   0.063    |  16.479 |\n",
      "| is_office                          |    0.823    |   0.140    |  5.893  |\n",
      "| is_retail                          |    0.446    |   0.129    |  3.454  |\n",
      "| jobs10                             |    1.137    |   0.051    |  22.455 |\n",
      "| np.log1p(jobs_within_30_min)       |    -0.346   |   0.084    |  -4.099 |\n",
      "| np.log1p(population_within_15_min) |    0.352    |   0.076    |  4.634  |\n",
      "| nonres_rent_per_sqft               |    0.115    |   0.011    |  10.765 |\n",
      "| is_peripheral                      |    -0.798   |   0.549    |  -1.452 |\n",
      "+------------------------------------+-------------+------------+---------+\n",
      "LCM RESULTS FOR SEGMENT 11\n",
      "\n",
      "Null Log-liklihood: -4605.170\n",
      "Log-liklihood at convergence: -1669.945\n",
      "Log-liklihood Ratio: 0.637\n",
      "\n",
      "+------------------------------------+-------------+------------+---------+\n",
      "| Component                          | Coefficient | Std. Error | T-Score |\n",
      "+------------------------------------+-------------+------------+---------+\n",
      "| jobs_800m                          |    0.043    |   0.044    |  0.963  |\n",
      "| ave_parcel_size                    |    1.225    |   0.061    |  19.969 |\n",
      "| is_office                          |    0.854    |   0.122    |  7.009  |\n",
      "| is_retail                          |    0.361    |   0.103    |  3.489  |\n",
      "| jobs11                             |    1.102    |   0.045    |  24.345 |\n",
      "| np.log1p(jobs_within_30_min)       |    -0.334   |   0.075    |  -4.471 |\n",
      "| np.log1p(population_within_15_min) |    0.450    |   0.068    |  6.580  |\n",
      "| nonres_rent_per_sqft               |    0.111    |   0.009    |  12.116 |\n",
      "| is_peripheral                      |    -3.000   |   0.733    |  -4.094 |\n",
      "+------------------------------------+-------------+------------+---------+\n",
      "LCM RESULTS FOR SEGMENT 12\n",
      "\n",
      "Null Log-liklihood: -4605.170\n",
      "Log-liklihood at convergence: -1168.723\n",
      "Log-liklihood Ratio: 0.746\n",
      "\n",
      "+------------------------------------+-------------+------------+---------+\n",
      "| Component                          | Coefficient | Std. Error | T-Score |\n",
      "+------------------------------------+-------------+------------+---------+\n",
      "| jobs_800m                          |    0.174    |   0.070    |  2.499  |\n",
      "| ave_parcel_size                    |    1.146    |   0.070    |  16.491 |\n",
      "| is_office                          |    0.419    |   0.132    |  3.166  |\n",
      "| is_retail                          |    -0.308   |   0.137    |  -2.257 |\n",
      "| jobs12                             |    0.888    |   0.056    |  15.806 |\n",
      "| np.log1p(jobs_within_30_min)       |    -0.068   |   0.083    |  -0.818 |\n",
      "| np.log1p(population_within_15_min) |    -0.140   |   0.080    |  -1.751 |\n",
      "| nonres_rent_per_sqft               |    0.108    |   0.010    |  11.307 |\n",
      "| is_peripheral                      |    -3.000   |   1.661    |  -1.806 |\n",
      "+------------------------------------+-------------+------------+---------+\n",
      "LCM RESULTS FOR SEGMENT 13\n",
      "\n",
      "Null Log-liklihood: -4605.170\n",
      "Log-liklihood at convergence: -570.676\n",
      "Log-liklihood Ratio: 0.876\n",
      "\n",
      "+------------------------------------+-------------+------------+---------+\n",
      "| Component                          | Coefficient | Std. Error | T-Score |\n",
      "+------------------------------------+-------------+------------+---------+\n",
      "| jobs_800m                          |    -0.132   |   0.074    |  -1.782 |\n",
      "| ave_parcel_size                    |    1.054    |   0.098    |  10.729 |\n",
      "| is_office                          |    0.318    |   0.201    |  1.581  |\n",
      "| is_retail                          |    -0.386   |   0.207    |  -1.863 |\n",
      "| jobs13                             |    1.451    |   0.062    |  23.481 |\n",
      "| np.log1p(jobs_within_30_min)       |    -0.241   |   0.107    |  -2.254 |\n",
      "| np.log1p(population_within_15_min) |    0.412    |   0.094    |  4.365  |\n",
      "| nonres_rent_per_sqft               |    0.129    |   0.015    |  8.631  |\n",
      "| is_peripheral                      |    -3.000   |   2.916    |  -1.029 |\n",
      "+------------------------------------+-------------+------------+---------+\n",
      "LCM RESULTS FOR SEGMENT 14\n",
      "\n",
      "Null Log-liklihood: -4605.170\n",
      "Log-liklihood at convergence: -1433.894\n",
      "Log-liklihood Ratio: 0.689\n",
      "\n",
      "+------------------------------------+-------------+------------+---------+\n",
      "| Component                          | Coefficient | Std. Error | T-Score |\n",
      "+------------------------------------+-------------+------------+---------+\n",
      "| jobs_800m                          |    0.031    |   0.054    |  0.577  |\n",
      "| ave_parcel_size                    |    1.302    |   0.067    |  19.346 |\n",
      "| is_office                          |    0.358    |   0.125    |  2.852  |\n",
      "| is_retail                          |    -0.264   |   0.109    |  -2.434 |\n",
      "| jobs14                             |    1.209    |   0.049    |  24.624 |\n",
      "| np.log1p(jobs_within_30_min)       |    -0.486   |   0.078    |  -6.212 |\n",
      "| np.log1p(population_within_15_min) |    0.438    |   0.072    |  6.104  |\n",
      "| nonres_rent_per_sqft               |    0.119    |   0.007    |  16.757 |\n",
      "| is_peripheral                      |    -3.000   |   1.921    |  -1.562 |\n",
      "+------------------------------------+-------------+------------+---------+\n",
      "LCM RESULTS FOR SEGMENT 15\n",
      "\n",
      "Null Log-liklihood: -4605.170\n",
      "Log-liklihood at convergence: -1415.848\n",
      "Log-liklihood Ratio: 0.693\n",
      "\n",
      "+------------------------------------+-------------+------------+---------+\n",
      "| Component                          | Coefficient | Std. Error | T-Score |\n",
      "+------------------------------------+-------------+------------+---------+\n",
      "| jobs_800m                          |    0.066    |   0.042    |  1.573  |\n",
      "| ave_parcel_size                    |    0.963    |   0.065    |  14.929 |\n",
      "| is_office                          |    -0.689   |   0.158    |  -4.353 |\n",
      "| is_retail                          |    -1.537   |   0.145    | -10.576 |\n",
      "| jobs15                             |    1.070    |   0.040    |  26.648 |\n",
      "| np.log1p(jobs_within_30_min)       |    -0.133   |   0.084    |  -1.584 |\n",
      "| np.log1p(population_within_15_min) |    0.145    |   0.078    |  1.853  |\n",
      "| nonres_rent_per_sqft               |    0.174    |   0.010    |  16.866 |\n",
      "| is_peripheral                      |    2.241    |   0.391    |  5.727  |\n",
      "+------------------------------------+-------------+------------+---------+\n",
      "LCM RESULTS FOR SEGMENT 16\n",
      "\n",
      "Null Log-liklihood: -4605.170\n",
      "Log-liklihood at convergence: -1509.628\n",
      "Log-liklihood Ratio: 0.672\n",
      "\n",
      "+------------------------------------+-------------+------------+---------+\n",
      "| Component                          | Coefficient | Std. Error | T-Score |\n",
      "+------------------------------------+-------------+------------+---------+\n",
      "| jobs_800m                          |    0.099    |   0.048    |  2.052  |\n",
      "| ave_parcel_size                    |    0.942    |   0.070    |  13.526 |\n",
      "| is_office                          |    0.785    |   0.120    |  6.566  |\n",
      "| is_retail                          |    -0.135   |   0.115    |  -1.172 |\n",
      "| jobs16                             |    0.972    |   0.041    |  23.568 |\n",
      "| np.log1p(jobs_within_30_min)       |    -0.338   |   0.082    |  -4.118 |\n",
      "| np.log1p(population_within_15_min) |    0.422    |   0.076    |  5.561  |\n",
      "| nonres_rent_per_sqft               |    0.137    |   0.011    |  12.884 |\n",
      "| is_peripheral                      |    -3.000   |   13.882   |  -0.216 |\n",
      "+------------------------------------+-------------+------------+---------+\n",
      "LCM RESULTS FOR SEGMENT 17\n",
      "\n",
      "Null Log-liklihood: -4605.170\n",
      "Log-liklihood at convergence: -1012.724\n",
      "Log-liklihood Ratio: 0.780\n",
      "\n",
      "+------------------------------------+-------------+------------+---------+\n",
      "| Component                          | Coefficient | Std. Error | T-Score |\n",
      "+------------------------------------+-------------+------------+---------+\n",
      "| jobs_800m                          |    -0.137   |   0.042    |  -3.274 |\n",
      "| ave_parcel_size                    |    0.954    |   0.067    |  14.167 |\n",
      "| is_office                          |    -0.061   |   0.165    |  -0.372 |\n",
      "| is_retail                          |    0.079    |   0.126    |  0.623  |\n",
      "| jobs17                             |    1.551    |   0.047    |  32.804 |\n",
      "| np.log1p(jobs_within_30_min)       |    -0.247   |   0.091    |  -2.723 |\n",
      "| np.log1p(population_within_15_min) |    0.155    |   0.084    |  1.838  |\n",
      "| nonres_rent_per_sqft               |    0.141    |   0.010    |  13.868 |\n",
      "| is_peripheral                      |    -0.502   |   0.691    |  -0.727 |\n",
      "+------------------------------------+-------------+------------+---------+\n",
      "LCM RESULTS FOR SEGMENT 18\n",
      "\n",
      "Null Log-liklihood: -4605.170\n",
      "Log-liklihood at convergence: -1630.639\n",
      "Log-liklihood Ratio: 0.646\n",
      "\n",
      "+------------------------------------+-------------+------------+---------+\n",
      "| Component                          | Coefficient | Std. Error | T-Score |\n",
      "+------------------------------------+-------------+------------+---------+\n",
      "| jobs_800m                          |    0.059    |   0.056    |  1.054  |\n",
      "| ave_parcel_size                    |    1.272    |   0.062    |  20.688 |\n",
      "| is_office                          |    0.416    |   0.131    |  3.181  |\n",
      "| is_retail                          |    0.983    |   0.092    |  10.633 |\n",
      "| jobs18                             |    0.908    |   0.053    |  17.193 |\n",
      "| np.log1p(jobs_within_30_min)       |    -0.338   |   0.067    |  -5.025 |\n",
      "| np.log1p(population_within_15_min) |    0.375    |   0.063    |  5.920  |\n",
      "| nonres_rent_per_sqft               |    0.121    |   0.007    |  16.734 |\n",
      "| is_peripheral                      |    -0.169   |   0.579    |  -0.292 |\n",
      "+------------------------------------+-------------+------------+---------+\n",
      "LCM RESULTS FOR SEGMENT 19\n",
      "\n",
      "Null Log-liklihood: -4605.170\n",
      "Log-liklihood at convergence: -2283.360\n",
      "Log-liklihood Ratio: 0.504\n",
      "\n",
      "+------------------------------------+-------------+------------+---------+\n",
      "| Component                          | Coefficient | Std. Error | T-Score |\n",
      "+------------------------------------+-------------+------------+---------+\n",
      "| jobs_800m                          |    0.050    |   0.037    |  1.352  |\n",
      "| ave_parcel_size                    |    1.090    |   0.064    |  17.089 |\n",
      "| is_office                          |    0.461    |   0.117    |  3.921  |\n",
      "| is_retail                          |    0.467    |   0.083    |  5.652  |\n",
      "| jobs19                             |    0.819    |   0.046    |  17.745 |\n",
      "| np.log1p(jobs_within_30_min)       |    -0.195   |   0.064    |  -3.031 |\n",
      "| np.log1p(population_within_15_min) |    0.524    |   0.061    |  8.555  |\n",
      "| nonres_rent_per_sqft               |    0.140    |   0.006    |  23.585 |\n",
      "| is_peripheral                      |    -2.887   |   0.882    |  -3.275 |\n",
      "+------------------------------------+-------------+------------+---------+\n",
      "LCM RESULTS FOR SEGMENT 20\n",
      "\n",
      "Null Log-liklihood: -4605.170\n",
      "Log-liklihood at convergence: -826.690\n",
      "Log-liklihood Ratio: 0.820\n",
      "\n",
      "+------------------------------------+-------------+------------+---------+\n",
      "| Component                          | Coefficient | Std. Error | T-Score |\n",
      "+------------------------------------+-------------+------------+---------+\n",
      "| jobs_800m                          |    0.134    |   0.056    |  2.401  |\n",
      "| ave_parcel_size                    |    1.021    |   0.080    |  12.746 |\n",
      "| is_office                          |    -1.129   |   0.176    |  -6.408 |\n",
      "| is_retail                          |    -1.370   |   0.147    |  -9.325 |\n",
      "| jobs20                             |    1.154    |   0.036    |  32.243 |\n",
      "| np.log1p(jobs_within_30_min)       |    -0.317   |   0.114    |  -2.781 |\n",
      "| np.log1p(population_within_15_min) |    0.153    |   0.104    |  1.476  |\n",
      "| nonres_rent_per_sqft               |    0.176    |   0.012    |  15.235 |\n",
      "| is_peripheral                      |    -0.250   |   0.545    |  -0.459 |\n",
      "+------------------------------------+-------------+------------+---------+\n",
      "Time to execute step 'elcm_estimate': 31.38 s\n",
      "Total time to execute iteration 1 with iteration value None: 31.39 s\n"
     ]
    }
   ],
   "source": [
    "orca.run([\"nrh_simulate\"])\n",
    "orca.run([\"nrh_simulate2\"])\n",
    "orca.run([\"elcm_estimate\"])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Running step 'jobs_transition'\n",
      "800800 agents before transition\n",
      "808808 agents after transition\n",
      "Time to execute step 'jobs_transition': 0.28 s\n",
      "Running step 'elcm_simulate'\n",
      "There are 989901 total available units\n",
      "    and 808808 total choosers\n",
      "    but there are 23348 overfull buildings\n",
      "    for a total of 626271 temporarily empty units\n",
      "    in 73074 buildings total in the region\n",
      "There are 8008 total movers for this LCM\n",
      "Assigned 8008 choosers to new units\n",
      "Total currently unplaced: 0\n",
      "    and there are now 181093 empty units\n",
      "    and 23347 overfull buildings\n",
      "Time to execute step 'elcm_simulate': 13.19 s\n",
      "Total time to execute iteration 1 with iteration value None: 13.48 s\n"
     ]
    }
   ],
   "source": [
    "orca.run([\"jobs_transition\", \"elcm_simulate\"])"
   ]
  },
  {
   "cell_type": "markdown",
   "metadata": {},
   "source": [
    "## HLCM Estimation"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 26,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Running step 'rsh_simulate'\n",
      "count    632386.000000\n",
      "mean        198.673383\n",
      "std         110.583603\n",
      "min           1.193853\n",
      "25%         133.371884\n",
      "50%         182.816569\n",
      "75%         245.252267\n",
      "max        3905.574146\n",
      "dtype: float64\n",
      "Time to execute step 'rsh_simulate': 13.52 s\n",
      "Total time to execute iteration 1 with iteration value None: 13.52 s\n",
      "Running step 'hlcm_estimate'\n",
      "LCM RESULTS FOR SEGMENT 1\n",
      "\n",
      "Null Log-liklihood: -27384.161\n",
      "Log-liklihood at convergence: -19426.715\n",
      "Log-liklihood Ratio: 0.291\n",
      "\n",
      "+------------------------------------+-------------+------------+---------+\n",
      "| Component                          | Coefficient | Std. Error | T-Score |\n",
      "+------------------------------------+-------------+------------+---------+\n",
      "| np.log1p(sqft_per_unit)            |    0.042    |   0.009    |  4.447  |\n",
      "| year_built                         |    -0.000   |   0.000    |  -1.742 |\n",
      "| np.log1p(residential_units)        |    1.232    |   0.014    |  90.443 |\n",
      "| jobs_800m                          |    0.029    |   0.008    |  3.654  |\n",
      "| residential_units_3000m            |    0.000    |   0.000    |  9.362  |\n",
      "| ave_income                         |    -0.166   |   0.042    |  -3.919 |\n",
      "| np.log1p(jobs_within_30_min)       |    -0.091   |   0.026    |  -3.455 |\n",
      "| np.log1p(population_within_15_min) |    0.198    |   0.030    |  6.718  |\n",
      "| res_price_per_sqft                 |    0.000    |   0.000    |  1.181  |\n",
      "| is_peripheral                      |    0.554    |   0.160    |  3.456  |\n",
      "+------------------------------------+-------------+------------+---------+\n",
      "LCM RESULTS FOR SEGMENT 2\n",
      "\n",
      "Null Log-liklihood: -27384.161\n",
      "Log-liklihood at convergence: -19731.804\n",
      "Log-liklihood Ratio: 0.279\n",
      "\n",
      "+------------------------------------+-------------+------------+---------+\n",
      "| Component                          | Coefficient | Std. Error | T-Score |\n",
      "+------------------------------------+-------------+------------+---------+\n",
      "| np.log1p(sqft_per_unit)            |    0.082    |   0.010    |  8.426  |\n",
      "| year_built                         |    -0.000   |   0.000    |  -0.503 |\n",
      "| np.log1p(residential_units)        |    1.260    |   0.014    |  91.431 |\n",
      "| jobs_800m                          |    0.013    |   0.008    |  1.658  |\n",
      "| residential_units_3000m            |    0.000    |   0.000    |  5.274  |\n",
      "| ave_income                         |    -0.109   |   0.040    |  -2.707 |\n",
      "| np.log1p(jobs_within_30_min)       |    -0.105   |   0.026    |  -4.066 |\n",
      "| np.log1p(population_within_15_min) |    0.162    |   0.029    |  5.505  |\n",
      "| res_price_per_sqft                 |    0.000    |   0.000    |  0.285  |\n",
      "| is_peripheral                      |    -0.204   |   0.178    |  -1.148 |\n",
      "+------------------------------------+-------------+------------+---------+\n",
      "LCM RESULTS FOR SEGMENT 3\n",
      "\n",
      "Null Log-liklihood: -27384.161\n",
      "Log-liklihood at convergence: -21083.682\n",
      "Log-liklihood Ratio: 0.230\n",
      "\n",
      "+------------------------------------+-------------+------------+---------+\n",
      "| Component                          | Coefficient | Std. Error | T-Score |\n",
      "+------------------------------------+-------------+------------+---------+\n",
      "| np.log1p(sqft_per_unit)            |    0.054    |   0.009    |  6.225  |\n",
      "| year_built                         |    0.001    |   0.000    |  1.451  |\n",
      "| np.log1p(residential_units)        |    1.200    |   0.014    |  86.210 |\n",
      "| jobs_800m                          |    -0.030   |   0.008    |  -3.831 |\n",
      "| residential_units_3000m            |    0.000    |   0.000    |  1.139  |\n",
      "| ave_income                         |    0.240    |   0.060    |  3.967  |\n",
      "| np.log1p(jobs_within_30_min)       |    -0.050   |   0.025    |  -1.973 |\n",
      "| np.log1p(population_within_15_min) |    0.090    |   0.029    |  3.109  |\n",
      "| res_price_per_sqft                 |    -0.000   |   0.000    |  -1.066 |\n",
      "| is_peripheral                      |    -0.155   |   0.164    |  -0.944 |\n",
      "+------------------------------------+-------------+------------+---------+\n",
      "LCM RESULTS FOR SEGMENT 4\n",
      "\n",
      "Null Log-liklihood: -27384.161\n",
      "Log-liklihood at convergence: -20760.855\n",
      "Log-liklihood Ratio: 0.242\n",
      "\n",
      "+------------------------------------+-------------+------------+---------+\n",
      "| Component                          | Coefficient | Std. Error | T-Score |\n",
      "+------------------------------------+-------------+------------+---------+\n",
      "| np.log1p(sqft_per_unit)            |    0.055    |   0.010    |  5.302  |\n",
      "| year_built                         |    0.000    |   0.000    |  0.576  |\n",
      "| np.log1p(residential_units)        |    1.232    |   0.014    |  86.353 |\n",
      "| jobs_800m                          |    -0.019   |   0.008    |  -2.459 |\n",
      "| residential_units_3000m            |    -0.000   |   0.000    |  -0.414 |\n",
      "| ave_income                         |    1.528    |   0.061    |  25.098 |\n",
      "| np.log1p(jobs_within_30_min)       |    -0.002   |   0.024    |  -0.087 |\n",
      "| np.log1p(population_within_15_min) |    0.055    |   0.030    |  1.868  |\n",
      "| res_price_per_sqft                 |    -0.000   |   0.000    |  -0.186 |\n",
      "| is_peripheral                      |    -0.126   |   0.184    |  -0.686 |\n",
      "+------------------------------------+-------------+------------+---------+\n",
      "Time to execute step 'hlcm_estimate': 264.89 s\n",
      "Total time to execute iteration 1 with iteration value None: 264.89 s\n"
     ]
    }
   ],
   "source": [
    "orca.run([\"rsh_simulate\"])\n",
    "orca.run([\"hlcm_estimate\"])"
   ]
  },
  {
   "cell_type": "code",
   "execution_count": 24,
   "metadata": {
    "collapsed": false
   },
   "outputs": [
    {
     "name": "stdout",
     "output_type": "stream",
     "text": [
      "Running iteration 1 with iteration value 2013\n",
      "Running step 'households_transition_basic'\n",
      "1034926 agents before transition\n",
      "1045275 agents after transition\n",
      "Time to execute step 'households_transition_basic': 0.55 s\n",
      "Total time to execute iteration 1 with iteration value 2013: 0.55 s\n",
      "Running step 'hlcm_simulate'\n",
      "There are 1161992 total available units\n",
      "    and 1045275 total choosers\n",
      "    but there are 1 overfull buildings\n",
      "    for a total of 127066 temporarily empty units\n",
      "    in 77497 buildings total in the region\n",
      "There are 10349 total movers for this LCM\n",
      "Assigned 10349 choosers to new units\n",
      "Total currently unplaced: 0\n",
      "    and there are now 116717 empty units\n",
      "    and 0 overfull buildings\n",
      "Time to execute step 'hlcm_simulate': 9.63 s\n",
      "Total time to execute iteration 1 with iteration value None: 9.64 s\n"
     ]
    }
   ],
   "source": [
    "orca.run([\"households_transition_basic\"], iter_vars = [2013])\n",
    "orca.run([\"hlcm_simulate\"])"
   ]
  }
 ],
 "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": "2.7.9"
  }
 },
 "nbformat": 4,
 "nbformat_minor": 0
}