example 05_glm_second_level uses fetch_localizer_contrasts() with option get_tmaps=True but tmaps aren’t used in this example.
What is the recommended way to modify this example? Open issue, fork, edit, open PR? Is an issue required?
You referring to these 2 examples: Second-level fMRI model: true positive proportion in clusters - Nilearn and Second-level fMRI model: two-sample test, unpaired and paired - Nilearn right?
You can open an issue first as a documentation issue so that we can track it and then open a pull request to resolve it. Please also see Contributing - Nilearn.
Let us know if you need any help at any point.