Updates for data var rename

Author: eric-czech

sgkit/stats/regenie.py

@@ -841,7 +841,7 @@ def regenie(
     G = ds[dosage]
     X = da.asarray(concat_2d(ds[list(covariates)], dims=("samples", "covariates")))
     Y = da.asarray(concat_2d(ds[list(traits)], dims=("samples", "traits")))
-    contigs = ds["variant/contig"]
+    contigs = ds["variant_contig"]
     return regenie_transform(
         G.T,
         X,

sgkit/tests/test_regenie.py

@@ -29,7 +29,7 @@
 from sgkit.testing import simulate_genotype_call_dataset
 
 regenie_sim = functools.partial(
-    regenie, dosage="call/dosage", covariates="sample/covariate", traits="sample/trait"
+    regenie, dosage="call_dosage", covariates="sample_covariate", traits="sample_trait"
 )
 
 
@@ -46,16 +46,16 @@ def simulate_regression_dataset(
     ds = simulate_genotype_call_dataset(
         n_variant=n_variant, n_sample=n_sample, n_contig=n_contig
     )
-    G = ds["call/genotype"].sum(dim="ploidy")
+    G = ds["call_genotype"].sum(dim="ploidy")
     X = rs.normal(size=(n_sample, n_covariate))
     Y = (
         G.T.data @ rs.normal(size=(G.shape[0], n_trait))
         + X @ rs.normal(size=(n_covariate, n_trait))
         + rs.normal(size=(n_sample, 1), scale=noise_scale)
     )
-    ds["call/dosage"] = G
-    ds["sample/covariate"] = (("samples", "covariates"), X)
-    ds["sample/trait"] = (("samples", "traits"), Y)
+    ds["call_dosage"] = G
+    ds["sample_covariate"] = (("samples", "covariates"), X)
+    ds["sample_trait"] = (("samples", "traits"), Y)
     return ds
 
 
@@ -185,7 +185,7 @@ def prepare_stage_3_sgkit_results(
             for k, v in dataclasses.asdict(stats).items()
         }
     )
-    dsr = dsr.merge(ds[["variant/id"]].rename({"variant/id": "variant_id"}))
+    dsr = dsr.merge(ds[["variant_id"]])
     dsr = dsr.assign(outcome=xr.DataArray(df_trait.columns, dims=("outcomes")))
     df = dsr.to_dataframe().reset_index(drop=True)  # type: ignore[no-untyped-call]
     return df
@@ -244,12 +244,10 @@ def check_simulation_result(
     # Load simulated data
     with zarr.ZipStore(str(dataset_dir / "genotypes.zarr.zip"), mode="r") as store:
         ds = xr.open_zarr(store)  # type: ignore[no-untyped-call]
-        # Temporary workaround for https://github.com/pystatgen/sgkit/issues/62
-        ds = ds.rename_vars({v: v.replace("-", "/") for v in ds})
         df_covariate = load_covariates(dataset_dir)
         df_trait = load_traits(dataset_dir)
-        contigs = ds["variant/contig"].values
-        G = ds["call/genotype"].sum(dim="ploidy").values
+        contigs = ds["variant_contig"].values
+        G = ds["call_genotype"].sum(dim="ploidy").values
         X = df_covariate.values
         Y = df_trait.values
 
@@ -308,7 +306,7 @@ def test_regenie__32bit_float(ds):
     ds = ds.assign(
         {
             v: ds[v].astype(np.float32)
-            for v in ["call/dosage", "sample/covariate", "sample/trait"]
+            for v in ["call_dosage", "sample_covariate", "sample_trait"]
         }
     )
     # Ensure that a uniform demotion in types for input arrays (aside from contigs)

sgkit/tests/test_regenie/dataset/sim_sm_01/genotypes.zarr.zip

@@ -26,4 +26,5 @@ invoke export --runs sim_sm_02-wgr_02 --runs sim_sm_01-wgr_01
 
 ### Glow WGR Release
 
-This validation was run for [glow.py==0.5.0](https://pypi.org/project/glow.py/0.5.0/).  At this time, binary traits are not yet supported and the REGENIE implementation hasn't even been officially released.  Support for [binary traits should come in the next release](https://github.com/projectglow/glow/issues/256) along with official support at which time this validation should be updated.  From that point onward, there is little need to update this data unless either implementation (sgkit or Glow) has been shown to be incorrect.
+This validation was run for [glow.py==0.5.0](https://pypi.org/project/glow.py/0.5.0/).  At this time, binary traits are not yet supported and the REGENIE implementation hasn't even been officially released.  Support for [binary traits should come in the next release](https://github.com/projectglow/glow/issues/256) along with official support at which time this validation should be updated.  From that point onward, there is little need to update this data unless either implementation (sgkit or Glow) has been shown to be incorrect.
+

sgkit/tests/test_regenie/dataset/sim_sm_02/genotypes.zarr.zip

@@ -6,6 +6,7 @@
 
 import io
 import logging
+import logging.config
 import shutil
 from pathlib import Path
 from typing import Dict, List, Optional

validation/gwas/method/regenie/README.md

@@ -2,6 +2,7 @@
 # coding: utf-8
 import io
 import logging
+import logging.config
 import shutil
 from pathlib import Path
 

validation/gwas/method/regenie/glow_wgr.py

@@ -2,6 +2,7 @@
 # coding: utf-8
 
 import logging
+import logging.config
 from pathlib import Path
 
 import fire
@@ -17,9 +18,7 @@ def run(dataset: str, dataset_dir="data/dataset"):
     dataset_dir = Path(dataset_dir)
     plink_path = dataset_dir / dataset / "genotypes"
     zarr_path = dataset_dir / dataset / "genotypes.zarr.zip"
-    ds = read_plink(plink_path, bim_sep="\t", fam_sep="\t")
-    # Temporary workaround for https://github.com/pystatgen/sgkit/issues/62
-    ds = ds.rename_vars({v: v.replace("/", "-") for v in ds})
+    ds = read_plink(path=plink_path, bim_sep="\t", fam_sep="\t")
     # Pre-compute string lengths until this is done:
     # https://github.com/pystatgen/sgkit-plink/issues/12
     ds = ds.compute()

validation/gwas/method/regenie/hail_sim.py

@@ -1,5 +1,6 @@
 import glob
 import logging
+import logging.config
 import os
 import shutil
 from pathlib import Path
@@ -73,7 +74,7 @@ def copy_files(src, dst, patterns):
     dst.mkdir(parents=True, exist_ok=True)
     files = [Path(f) for pattern in patterns for f in glob.glob(str(src / pattern))]
     for f in files:
-        logger.info("\tCopying path: {f}")
+        logger.info(f"\tCopying path: {f}")
         if f.is_dir():
             shutil.copytree(f, dst / f.name)
         else: