obs_var
- pyoptex.doe.cost_optimal.utils.obs_var(Yenc, colstart, ratios=None, grouped_cols=None)[source]
Directly computes the observation matrix from the design. Is similar to
obs_var_Zsfollowed byobs_var_from_Zs.Parameters
- Yencnp.array(2d)
The categorically encoded design matrix.
- colstartnp.array(1d)
The start column of each factor.
- ratiosnp.array(1d)
The variance ratios of the different groups compared to the variance of the random errors.
- grouped_colsnp.array(1d)
A boolean array indicating whether the factor is grouped or not.
Returns
- Vnp.array(2d)
The observation covariance matrix.