I am trying to download a repository using datalad clone command. It downloaded the dataset with symlinks. So I used datalad get to get the actual content of file. Though these commands run without any problems, I still cannot open the file. It’s showing symlink files:
What are you trying to open the files with, and how are you trying to open the file? How much storage do you see a downloaded .nii.gz taking on your filesystem?
As those are .gz file, I simply tried to unzip it. So, for this file (sub-S001_T1w_defaced.nii.gz) I am expecting a NIFTI file with similar name, instead it’s creating MD5E-s10835011–34bcbedf42ad11a3f98d25aec0a0a530.nii.
The size of downloaded file is 148 bytes whereas the actual size is 10MB.
Can you share your datalad get command and any terminal outputs when running it? To clarify, what you should be doing is something like
cd $DATALAD_DATASET_ROOT
### You can also cd further to the folder you want to get data from, if there's a specific one
datalad get $FILENAME
### If you are not in the folder where the file is, you need to provide a path to that file
### Wildcard patterns e.g., sub-01*.nii.gz, are acceptable
This suggests that the file did not download, as 148 bytes sounds like the size of an empty symlink.
That looks good. Keep in mind the data in the repo are still symlinks, but now should be linked to a bigger file. What were you trying to open the nifti with, and does it work now?
There’s a weird GIN/git-annex quirk where using the .git suffix disables the annex remote. You should be able to git annex enableremote origin, but if that doesn’t work, re-clone without the .git.
I tried to run enableremote command but it failed with this message:
enableremote (normal) origin
Remote origin not usable by git-annex; setting annex-ignore
https://gin.g-node.org/psytest/TSYD_0101/config download failed: Not Found
failed
enableremote: 1 failed
I also clone the repo without .git (datalad clone https://gin.g-node.org/psytest/TSYD_0101) but cannot download the actual file content. The direcory has only the symlink files.
The directory will always have symlinks. The linked files are where the data are contained. If you, for example, freeview one of the files, it should follow the symlink and load the data.
If it’s very important not to have symlinks, you can unlock the files:
datalad unlock .
Note that this will ~double the amount of space the dataset consumes.
I am trying to download the data from repo. I have tried using both the datalad clone and datalad install commands, and then running datalad get . -r to get the actual file content. However, I keep getting the error message get (error) [Not available; .git remotes have annex-ignore set:origin. Just to check that I am creating the repository correctly, here are the commands I used:
datalad create -c text2git TSYD_0101
cd TSYD_0101
datalad save -m "add files in repo"
datalad siblings add -d . \
--name gin \
--url git@gin.g-node.org:/psytest/TSYD_0101.git
datalad push --to gin
