Sorry I could not run it because I don't see how to get the ds107 dataset. Maybe this should be explained somewhere ? Should we get it from openfmri ? Or have you added it to nipy-data ? What data is expected actually ?

Hi Bertrand - sorry - I did not explain. I'm thinking that the ds105 dataset might be the best dataset to use. I guess we should get it from openfmri and I need to add instructions here.

btw, these datasets are now on amazon s3, so if you only wanted to download specific files, you can directly embed the urls for those files.

http://openfmri.aws.amazon.com.s3.amazonaws.com/ds105/... will get you files from ds105.

@satra - your url for the Amazon openfmri dataset seems to be broken - has it changed?

it hasn't quite changed but you need to know the path to the exact file. amazon doesn't allow getting at folder contents.

for example:

http://openfmri.aws.amazon.com.s3.amazonaws.com/ds005/README

will get you the file

temporarily you can get that path to append from:

https://openfmri.org/system/files/ds005_raw_checksums.txt

i'll get you a better way to query them together soon.

I adapted the example to ds105 instead.

I also adapted the code to use SPM8.

I added some instructions at the top of the example.

Changes Unknown when pulling 5b6fc82 on matthew-brett:open-fmri-spm-example into * on nipy:master*.

Any comments here?

@matthew-brett - looks good.

this might be a possible example where you could point to nipype for extended capabilities to interface with external tools and for dataflow management through the workflow engine. from the spm perspective, nipype can handle vanilla spm in matlab or spm run through the matlab common runtime (free from matlab licensing).

Comment in file OK?

thanks @matthew-brett - sounds good

Changes Unknown when pulling 26665b2 on matthew-brett:open-fmri-spm-example into * on nipy:master*.

OK to merge this one? Will merge tomorrow unless I hear otherwise.

+1