If I understand the code correctly, then the Nipy GLM produces non-standardized beta coefficients (i.e. they can be larger than 1 and smaller than -1). This should be stated somewhere in the documentation as beta usually refers to the standardized values.
If I understand the code correctly, then the Nipy GLM produces non-standardized beta coefficients (i.e. they can be larger than 1 and smaller than -1). This should be stated somewhere in the documentation as beta usually refers to the standardized values.