How to use Python StatsModels To estimate logit Return to .
consider a logit regression of employ on yoe, edu, unhealthy and cityhukou as if we are using a poole--d cross-sectional data. Specifically,
Pr(𝑒𝑚𝑝𝑙𝑜𝑦𝑖𝑡=1)=Λ(𝛽0+𝛽1𝑦𝑜𝑒𝑖𝑡+𝛽2𝑒𝑑𝑢𝑖𝑡+𝛽3𝑢𝑛ℎ𝑒𝑎𝑙𝑡ℎ𝑦𝑖𝑡+𝛽4𝑐𝑖𝑡𝑦ℎ𝑢𝑘𝑜𝑢𝑖𝑡)
where Λ(𝑧)=𝑒xp(𝑧)/[1+exp(𝑧)] , subscripts 𝑖 and 𝑡 are indices for cross-sections of respondents and years respectively, i.e., 𝑖=1,…,856;𝑡=2002,…,2013 . Please try to respectively use Python StatsModels and Python machine learning module (Scikit-learn) to estimate this model and summarize your findings.
