ValueError: data type must provide an itemsize

[alimanfoo]

This code run on colab:

sample_sets = ["AG1000G-MW"]
sample_query = "taxon = 'arabiensis'"
cyp6aap_region = "2R:28,480,000-28,510,000"
df_cyp6aap_cnv = ag3.gene_cnv_frequencies(
    region=cyp6aap_region,
    cohorts="admin2_year",
    sample_sets=sample_sets,
    sample_query=sample_query,
)
df_cyp6aap_cnv

...generates an error:

---------------------------------------------------------------------------
ValueError                                Traceback (most recent call last)
[<ipython-input-12-6f50beb47483>](https://localhost:8080/#) in <cell line: 0>()
      2 sample_query = "taxon = 'arabiensis'"
      3 cyp6aap_region = "2R:28,480,000-28,510,000"
----> 4 df_cyp6aap_cnv = ag3.gene_cnv_frequencies(
      5     region=cyp6aap_region,
      6     cohorts="admin2_year",

15 frames
[/usr/local/lib/python3.11/dist-packages/malariagen_data/util.py](https://localhost:8080/#) in check_types_wrapper(*args, **kwargs)
   1162                     error = TypeError(message)
   1163                     raise error from None
-> 1164         return f(*args, **kwargs)
   1165 
   1166     return check_types_wrapper

[/usr/local/lib/python3.11/dist-packages/malariagen_data/anoph/cnv_frq.py](https://localhost:8080/#) in gene_cnv_frequencies(***failed resolving arguments***)
    217         debug("access and concatenate data from regions")
    218         df = pd.concat(
--> 219             [
    220                 self._gene_cnv_frequencies(
    221                     region=r,

[/usr/local/lib/python3.11/dist-packages/malariagen_data/anoph/cnv_frq.py](https://localhost:8080/#) in <listcomp>(.0)
    218         df = pd.concat(
    219             [
--> 220                 self._gene_cnv_frequencies(
    221                     region=r,
    222                     cohorts=cohorts,

[/usr/local/lib/python3.11/dist-packages/malariagen_data/anoph/cnv_frq.py](https://localhost:8080/#) in _gene_cnv_frequencies(self, region, cohorts, sample_query, sample_query_options, min_cohort_size, sample_sets, drop_invariant, max_coverage_variance, include_counts, chunks, inline_array)
    263 
    264         debug("get gene copy number data")
--> 265         ds_cnv = self.gene_cnv(
    266             region=region,
    267             sample_sets=sample_sets,

[/usr/local/lib/python3.11/dist-packages/malariagen_data/util.py](https://localhost:8080/#) in check_types_wrapper(*args, **kwargs)
   1162                     error = TypeError(message)
   1163                     raise error from None
-> 1164         return f(*args, **kwargs)
   1165 
   1166     return check_types_wrapper

[/usr/local/lib/python3.11/dist-packages/malariagen_data/anoph/cnv_frq.py](https://localhost:8080/#) in gene_cnv(***failed resolving arguments***)
     61 
     62         ds = simple_xarray_concat(
---> 63             [
     64                 self._gene_cnv(
     65                     region=r,

[/usr/local/lib/python3.11/dist-packages/malariagen_data/anoph/cnv_frq.py](https://localhost:8080/#) in <listcomp>(.0)
     62         ds = simple_xarray_concat(
     63             [
---> 64                 self._gene_cnv(
     65                     region=r,
     66                     sample_sets=sample_sets,

[/usr/local/lib/python3.11/dist-packages/malariagen_data/anoph/cnv_frq.py](https://localhost:8080/#) in _gene_cnv(self, region, sample_sets, sample_query, sample_query_options, max_coverage_variance, chunks, inline_array)
    105 
    106         # Access HMM data.
--> 107         ds_hmm = self.cnv_hmm(
    108             region=cnv_region,
    109             sample_sets=sample_sets,

[/usr/local/lib/python3.11/dist-packages/malariagen_data/util.py](https://localhost:8080/#) in check_types_wrapper(*args, **kwargs)
   1162                     error = TypeError(message)
   1163                     raise error from None
-> 1164         return f(*args, **kwargs)
   1165 
   1166     return check_types_wrapper

[/usr/local/lib/python3.11/dist-packages/malariagen_data/anoph/cnv_data.py](https://localhost:8080/#) in cnv_hmm(***failed resolving arguments***)
    256                 debug("apply the query")
    257                 sample_query_options = sample_query_options or {}
--> 258                 loc_query_samples = df_samples_cnv.eval(
    259                     sample_query, **sample_query_options
    260                 ).values

[/usr/local/lib/python3.11/dist-packages/pandas/core/frame.py](https://localhost:8080/#) in eval(self, expr, inplace, **kwargs)
   4947         kwargs["resolvers"] = tuple(kwargs.get("resolvers", ())) + resolvers
   4948 
-> 4949         return _eval(expr, inplace=inplace, **kwargs)
   4950 
   4951     def select_dtypes(self, include=None, exclude=None) -> Self:

[/usr/local/lib/python3.11/dist-packages/pandas/core/computation/eval.py](https://localhost:8080/#) in eval(expr, parser, engine, local_dict, global_dict, resolvers, level, target, inplace)
    355         eng = ENGINES[engine]
    356         eng_inst = eng(parsed_expr)
--> 357         ret = eng_inst.evaluate()
    358 
    359         if parsed_expr.assigner is None:

[/usr/local/lib/python3.11/dist-packages/pandas/core/computation/engines.py](https://localhost:8080/#) in evaluate(self)
     79 
     80         # make sure no names in resolvers and locals/globals clash
---> 81         res = self._evaluate()
     82         return reconstruct_object(
     83             self.result_type, res, self.aligned_axes, self.expr.terms.return_type

[/usr/local/lib/python3.11/dist-packages/pandas/core/computation/engines.py](https://localhost:8080/#) in _evaluate(self)
    119         scope = env.full_scope
    120         _check_ne_builtin_clash(self.expr)
--> 121         return ne.evaluate(s, local_dict=scope)
    122 
    123 

[/usr/local/lib/python3.11/dist-packages/numexpr/necompiler.py](https://localhost:8080/#) in evaluate(ex, local_dict, global_dict, out, order, casting, sanitize, _frame_depth, **kwargs)
    971                  _frame_depth=_frame_depth, sanitize=sanitize, **kwargs)
    972     if e is None:
--> 973         return re_evaluate(local_dict=local_dict, global_dict=global_dict, _frame_depth=_frame_depth)
    974     else:
    975         raise e

[/usr/local/lib/python3.11/dist-packages/numexpr/necompiler.py](https://localhost:8080/#) in re_evaluate(local_dict, global_dict, _frame_depth)
   1003     kwargs = _numexpr_last['kwargs']
   1004     with evaluate_lock:
-> 1005         return compiled_ex(*args, **kwargs)

ValueError: data type must provide an itemsize

It's related to numexpr as the query engine to evaluate a pandas query, in this case on the sample metadata. Usually this is the default and works fine, but is now generating this error for some reason.

Here is a partial workaround:

sample_sets = ["AG1000G-MW"]
sample_query = "taxon = 'arabiensis'"
cyp6aap_region = "2R:28,480,000-28,510,000"
df_cyp6aap_cnv = ag3.gene_cnv_frequencies(
    region=cyp6aap_region,
    cohorts="admin2_year",
    sample_sets=sample_sets,
    sample_query=sample_query,
    sample_query_options=dict(engine="python"),
)
df_cyp6aap_cnv

...although that then generates a different exception (which I'll raise a separate issue about).

Versions: