Continuous vs categorical variables in CCA and post-hoc regress

bernard-liew commented 9 years ago

Dear Todd,

I would like to know if CCA and regress functions in spm1d treats the independent variables as continuous or categorically.

Regards, Bernard

0todd0000 commented 9 years ago

Hi Bernard, The answer is: continuous.

For regression:

independent variable: univariate 0D (continuous 0D scalar)
dependent variable: univariate 0D or 1D (continuous 0D or 1D scalar)

For CCA (as currently implemented in spm1d):

independent variable: univariate 0D (continuous 0D scalar)
dependent variable: multivariate 0D or 1D (continuous 0D or 1D vector)

In general, CCA can be used with multivariate independent variables, but that is not yet implemented in spm1d. As far as I know the terms "regression" and "CCA" are not used when the independent variables are categorical.

Todd

bernard-liew commented 9 years ago

Many thanks Todd:)

Kind regards, Bernard

0todd0000 / spm1d

Continuous vs categorical variables in CCA and post-hoc regress #27