opal-analytics-logit2

Installation

pip install git+https://github.com/Bedrock-py/opal-analytics-logit2.git

Parameters Spec for Bedrock

self.parameters_spec = [
    { "name" : "Regression Formula", "attrname" : "formula", "value" : "", "type" : "input" },
    { "name" : "GLM family", "attrname" : "family", "value" : "binomial", "type" : "input" },
    { "name" : "Clustered Error Covariates", "attrname" : "clustered_rse" , "value" : "", "type" : "input"}
]

• formula A R-style formula for regression given as a string
• family Either binomial for logistic regression or gaussian for OLS
• clustered_rse Columns that should be clustered for robust standard error.
  The format of the parameter is a comma delimited string "column1,column2"

Requires a Matrix with the following files

• matrix.csv The full matrix with both endogenous and exogenous variables
• features.txt A list of column names for the matrix (one name per row)

Outputs the following files

matrix.csv The coefficient table output from the GLM model summary.csv Summary metrics for the model fit