balance_preditctor.py 文件源码

def build_model(y, x, add_constant=True):
    '''
        Build a linear regression model from the provided data

        Provided:
        y: a series or single column dataframe holding our solution vector for linear regression
        x: a dataframe that runs parallel to y, with all the features for our linear regression
        add_constant: a boolean, if true it will add a constant row to our provided x data. Otherwise
                      this method assumes you've done-so already, or do not want one for some good reason

        Return: a linear regression model from StatsModels and the data which was used to train the model.
                If add_constant was true this will be a new dataframe, otherwise it will be x
    '''
    if add_constant:
        x = sm.add_constant(x)

    model = sm.OLS(y, x).fit()
    return (model, x)