gimble parse error on example data

[arzwa]

I am trying to run gIMble on the example data set included in the git repo. I did gimbleprep -f reference.fa -b bams/ -v heliconius.test_data.vcf.gz -k and then edited the resulting gimble.samples.csv file to add the population labels:

chi.CJ564.m,p1
chi.CJ565.m,p1
ros.CAM2045.m,p2
ros.CAM2059.m,p2
ros.CAM2552.m,p2

If I now run gimble parse -g gimble.genomefile -v heliconius.test_data.vcf.gz -b gimble.bed -s gimble.samples.csv -z heliconius.v1 -f I get the following error:

[+] Running 'gimble parse' ...
[-] Gimble datastore 'heliconius.v1.z' already exists.
[+] Deleting existing Gimble datastore 'heliconius.v1.z'
[+] Creating Gimble datastore in 'heliconius.v1.z'
[#] Processing GENOME_FILE '/home/arzwa/dev/gIMble/example_input/gimble.genomefile'.
[#] Processing SAMPLE_FILE '/home/arzwa/dev/gIMble/example_input/gimble.samples.csv'.
[#] Processing BED_FILE '/home/arzwa/dev/gIMble/example_input/gimble.bed'.
[%] Reading intervals:   0%|                                                  | 0/3 [00:00<?,[%] Reading intervals:   0%|                                                  | 0/3 [00:00<?, ?it/s]
Traceback (most recent call last):
  File "/home/arzwa/miniconda3/bin//gimble", line 10, in <module>
    sys.exit(main())
             ~~~~^^
  File "/home/arzwa/miniconda3/lib/python3.13/site-packages/cli/interface.py", line 95, in main
    runner.main(params)
    ~~~~~~~~~~~^^^^^^^^
  File "/home/arzwa/miniconda3/lib/python3.13/site-packages/cli/parse.py", line 54, in main
    gimbleStore.measure(
    ~~~~~~~~~~~~~~~~~~~^
        genome_f=parameterObj.genome_f,
        ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
        sample_f=parameterObj.sample_f,
        ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
        bed_f=parameterObj.bed_f,
        ^^^^^^^^^^^^^^^^^^^^^^^^^
        vcf_f=parameterObj.vcf_f)
        ^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/home/arzwa/miniconda3/lib/python3.13/site-packages/lib/gimble.py", line 1175, in measure
    self._read_intervals(measure_key, bed_f)
    ~~~~~~~~~~~~~~~~~~~~^^^^^^^^^^^^^^^^^^^^
  File "/home/arzwa/miniconda3/lib/python3.13/site-packages/lib/gimble.py", line 1386, in _read_intervals
    self._set_intervals(sequence, interval_matrix, starts, ends)
    ~~~~~~~~~~~~~~~~~~~^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
  File "/home/arzwa/miniconda3/lib/python3.13/site-packages/lib/gimble.py", line 1398, in _set_intervals
    self.data.create_dataset(
    ~~~~~~~~~~~~~~~~~~~~~~~~^
        "seqs/%s/intervals/matrix" % sequence, data=interval_matrix
        ^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
    )
    ^
  File "/home/arzwa/miniconda3/lib/python3.13/warnings.py", line 637, in wrapper
    return arg(*args, **kwargs)
  File "/home/arzwa/miniconda3/lib/python3.13/site-packages/zarr/core/group.py", line 2786, in create_dataset
    return Array(self._sync(self._async_group.create_dataset(name, **kwargs)))
                            ~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~~^^^^^^^^^^^^^^^^
  File "/home/arzwa/miniconda3/lib/python3.13/warnings.py", line 637, in wrapper
    return arg(*args, **kwargs)
TypeError: AsyncGroup.create_dataset() missing 1 required keyword-only argument: 'shape'

I am not a regular python/conda/... user, so I might have messed up something in installation. I did use conda though, as suggested in the README, and installation appeared to go smoothly.

Any ideas/suggestions?

[GertjanBisschop]

Thanks @arzwa for this bugreport. Looking at the zarr documentation, this most likely has to do with upstream API changes (zarr in this case) and a lack of version pinning within gimble of this zarr dependency. I've got no idea when someone would have the time to fix this. You could try and downgrade zarr yourself to either 2.* or 1.* depending on when the shape keyword was introduced.

[arzwa]

Pinning zarr==2.* did the trick, thanks!