[pymvpa] An error occurs when I use the function "zscore" ...

[Michael Hanke]

On Thu, Feb 09, 2012 at 07:30:06PM +0800, 冯璐 wrote:
> 
> But if I use the code like this,
> 	zscore(dataset, targetdtype="float32"),
> the result seems reasonable.
> 
> The declaration of the function "zscore" is like this,
> 	zscore(dataset, mean=None, std=None, perchunk=True, baselinelabels=None, pervoxel=True, targetdtype="float32").
> It also said that,
> 	"If 'baselinelabels' provided, and 'mean' or 'std' aren't provided, it would compute the corresponding measure based only on labels in 'baselinelabels'."
> 
> I'm puzzled. When mean=std=baselinelabels=None, what happened to the dataset?

In this case the features get zscored base on the feature mean and std (computed
over all samples). mean, std are there to allow specification of
particular values derived from something else than the dataset at hands.
baselinelabels allows you to derive the mean and std from a subset of
samples within the dataset -- maybe to compute something like "deviation
from baseline" by zscoring based on that.

HTH,

Michael

-- 
Michael Hanke
http://mih.voxindeserto.de