fit_mixedlm
- pyoptex.analysis.utils.fit.fit_mixedlm(X, y, groups)[source]
Wrapper to fit a statsmodels Mixed LM model. It includes the following additional attributes required for a more synchronized output between OLS and Mixed LM.
rsquared : The simple r2_score of the fit.
rsquared_adj : The adjust r2 score according to
r2_adj.
Parameters
- Xnp.array(2d)
The normalized, encoded model matrix of the data.
- ynp.array(1d)
The output variable
- groupsnp.array(2d)
The 2d array of the groups for the random effects of size (g, N), with g the number of random effects and N the number of runs. Each row is an integer array with values from 0 until the total number of groups for that random effect. For example [0, 0, 1, 1] indicates that the first two runs are correlated, and the last two.
Returns
- fit
statsmodels.regression.mixed_linear_model.MixedLMResults The statsmodels Mixed LM results with some additional attributes.