fix: resolve KeyError and fragile array slicing in haplotypes_frequencies_advanced()`

[blankirigaya]

Summary

Fixes KeyError: 'label' that makes haplotypes_frequencies_advanced()
completely unusable, and replaces fragile positional .to_numpy() slicing
with explicit named column selection.

Closes #<ISSUE_NUMBER>

Changes

In malariagen_data/anoph/hap_frq.py:

1. Remove premature set_index("label") call that was dropping the
   label column before it could be read
2. Replace df.to_numpy()[:, :n] positional slices with
   df[freq_cols].to_numpy() etc. using explicit column name lists
3. Drop the now-unnecessary set_index entirely (return value is
   ds_out, not df_haps_sorted)

# Before
df_haps_sorted["label"] = [...]
df_haps_sorted.set_index(keys="label", drop=True, inplace=True)
ds_out["variant_label"] = "variants", df_haps_sorted["label"]   # KeyError
ds_out["event_frequency"] = ..., df_haps_sorted.to_numpy()[:, :n]
ds_out["event_count"]     = ..., df_haps_sorted.to_numpy()[:, n:2*n]
ds_out["event_nobs"]      = ..., df_haps_sorted.to_numpy()[:, 2*n:-2]

# After
labels = ["H" + str(i) for i in range(len(df_haps_sorted))]
df_haps_sorted["label"] = labels
freq_cols  = [c for c in df_haps_sorted.columns if c.startswith("frq_")]
count_cols = [c for c in df_haps_sorted.columns if c.startswith("count_")]
nobs_cols  = [c for c in df_haps_sorted.columns if c.startswith("nobs_")]
ds_out["variant_label"]   = "variants", df_haps_sorted["label"].values
ds_out["event_frequency"] = ("variants", "cohorts"), df_haps_sorted[freq_cols].to_numpy()
ds_out["event_count"]     = ("variants", "cohorts"), df_haps_sorted[count_cols].to_numpy()
ds_out["event_nobs"]      = ("variants", "cohorts"), df_haps_sorted[nobs_cols].to_numpy()

Files Changed

• malariagen_data/anoph/hap_frq.py