Maps threshold bug

[BTajini]

Hi,

I believe there's a bug in the maps threshold.
As you can see in the following screenshot.

In the t_stats , we have some negative zones while in Raw z map, they are positive.

Best,

[bthirion]

Can you send me the code you used ? I can't reproduce the issue.

[bthirion]

OK, the problem comes from the fact that you obtained the p-values from a ttest_onesamp of scipy.stats that is a two-sided tests, and implicitly takes the absolute value.
You should use nistats functions, not ttest_onesamp, in that context (copy-pasting from old notebooks has some drawbacks !).
Best,