This repository contains the code for the microsimulation project
-
The
1_clean_FMLA.py
file takes thedata/fmla_2012/fmla_2012_employee_restrict_puf.csv
data as input, creates a number of variables required for the estimation procedures and saves thedata/fmla_clean_2012.csv
file. -
The
1a_get_response.py
file takes thedata/fmla_clean_2012.csv
data as input, creates a new column 'resp_length' to represent counterfactual leave length for FMLA workers in presence of a program -
(to be integrated later) The
2.estimate_behavioral.py
file taks thedata/fmla_clean_2012.csv
data as input and estimates a number of logit regressions. The resulting parameter estimates are saved as csv files in theestimates
folder. -
(to be integrated later) The
2a_estimate_behavioral_cps
file does XYZ... -
(to be integrated later) The
3.estimate_distributions.py
file taks thefmla_clean_2012.csv
data as input and calculates a number of leave length distributions, which is then saves as csv files in theestimates
folder. -
The
4_clean_ACS.py
file takes the raw ACS files in thedata/acs
repository and creates the necessary variables to be consistent with the output from the estimation component and saves a master analysis data file calledACS_cleaned_forsimulation.csv
. -
The
5_simulation_engine.py
file runs the simulation. -
The
5a_simulate_knn.py
file contains functions (knn) neeeded for simulation. -
(to be integrated later) The
5z_simulate_hk.py
file contains earilier simulation code by Hautahi Kingi.