Globbing should be advertised in doc/examples.

[AlexandreAbraham]

No description provided.

[lesteve]

It is already in doc/manipulating_visualizing/manipulating_images.rst. Mentioning where it could be added would help.

[aabadie]

It is already in doc/manipulating_visualizing/manipulating_images.rst. Mentioning where it could be added would help.

+1 ;)

[AlexandreAbraham]

@GaelVaroquaux, have you some specific location in mind?

[GaelVaroquaux]

It is already in doc/manipulating_visualizing/manipulating_images.rst.
Mentioning where it could be added would help.

Maybe in the "Getting started", where we talk about plotting, say, in a
sidebar.

We need to put ourselves in the shoes of the user. How does the user
discover this feature? I am not saying that it is easy to figure this
out, but we need to think about it.

[man-shu]

Reviving this discussion, I think this introductory example would be a good candidate.

As of now, the plotting is shown on a single file fetched from neurovault, but to glob we need to fetch several files, for example, with nilearn.datasets.fetch_mixed_gambles. So if this sounds good I could:

• swap the datasets.fetch_neurovault_motor_task with datasets.fetch_mixed_gambles for n subjects.
• show possibility of globbing over these subjects (other than the dictionary the fetcher already returns) with image.mean_img and plotting this mean stat map.

[Remi-Gau]

That sounds good though "a bit abstract" (sorry I have not had my second coffee yet), so maybe open a quick draft PR to show what you mean.

[man-shu]

Yes, on it!