Merge branch 'master' into 367-haps-freq

Author: jonbrenas

docs/source/Af1.rst

@@ -53,6 +53,8 @@ Sample metadata access
     count_samples
     plot_samples_bar
     plot_samples_interactive_map
+    plot_sample_location_mapbox
+    plot_sample_location_geo
     wgs_data_catalog
     cohorts
 
@@ -132,6 +134,7 @@ Genetic distance and neighbour-joining trees (NJT)
     :toctree: generated/
 
     plot_njt
+    njt
     biallelic_diplotype_pairwise_distances
 
 Heterozygosity analysis
@@ -161,6 +164,8 @@ Genome-wide selection scans
     plot_h12_calibration
     h12_gwss
     plot_h12_gwss
+    plot_h12_gwss_multi_panel
+    plot_h12_gwss_multi_overlay
     h1x_gwss
     plot_h1x_gwss
     g123_calibration

docs/source/Ag3.rst

@@ -54,6 +54,8 @@ Sample metadata access
     lookup_sample
     plot_samples_bar
     plot_samples_interactive_map
+    plot_sample_location_mapbox
+    plot_sample_location_geo
     wgs_data_catalog
     cohorts
 
@@ -142,6 +144,7 @@ Genetic distance and neighbour-joining trees (NJT)
     :toctree: generated/
 
     plot_njt
+    njt
     biallelic_diplotype_pairwise_distances
 
 Heterozygosity analysis
@@ -171,6 +174,8 @@ Genome-wide selection scans
     plot_h12_calibration
     h12_gwss
     plot_h12_gwss
+    plot_h12_gwss_multi_panel
+    plot_h12_gwss_multi_overlay
     h1x_gwss
     plot_h1x_gwss
     g123_calibration

malariagen_data/af1.py

@@ -40,7 +40,7 @@ class Af1(AnophelesDataResource):
     debug : bool, optional
         Set to True to enable debug level logging.
     show_progress : bool, optional
-        If True, show a progress bar during longer-running computations.
+        If True, show a progress bar during longer-running computations. The default can be overridden using an environmental variable named MGEN_SHOW_PROGRESS.
     check_location : bool, optional
         If True, use ipinfo to check the location of the client system.
     **kwargs
@@ -82,7 +82,7 @@ def __init__(
         results_cache=None,
         log=sys.stdout,
         debug=False,
-        show_progress=True,
+        show_progress=None,
         check_location=True,
         cohorts_analysis=None,
         site_filters_analysis=None,

malariagen_data/ag3.py

@@ -112,7 +112,7 @@ class Ag3(AnophelesDataResource):
     debug : bool, optional
         Set to True to enable debug level logging.
     show_progress : bool, optional
-        If True, show a progress bar during longer-running computations.
+        If True, show a progress bar during longer-running computations. The default can be overridden using an environmental variable named MGEN_SHOW_PROGRESS.
     check_location : bool, optional
         If True, use ipinfo to check the location of the client system.
     **kwargs
@@ -154,7 +154,7 @@ def __init__(
         results_cache=None,
         log=sys.stdout,
         debug=False,
-        show_progress=True,
+        show_progress=None,
         check_location=True,
         cohorts_analysis=None,
         aim_analysis=None,
@@ -378,6 +378,7 @@ def karyotype(
         inversion: inversion_param,
         sample_sets: Optional[base_params.sample_sets] = None,
         sample_query: Optional[base_params.sample_query] = None,
+        sample_query_options: Optional[base_params.sample_query_options] = None,
     ) -> pd.DataFrame:
         # load tag snp data
         df_tagsnps = self.load_inversion_tags(inversion=inversion)
@@ -390,19 +391,25 @@ def karyotype(
         region = f"{contig}:{start}-{end}"
 
         ds_snps = self.snp_calls(
-            region=region, sample_sets=sample_sets, sample_query=sample_query
+            region=region,
+            sample_sets=sample_sets,
+            sample_query=sample_query,
+            sample_query_options=sample_query_options,
         )
-        geno = allel.GenotypeDaskArray(ds_snps["call_genotype"].data)
-        pos = allel.SortedIndex(ds_snps["variant_position"].values)
-        samples = ds_snps["sample_id"].values
-        alts = ds_snps["variant_allele"].values.astype(str)
-
-        # subset to position of inversion tags
-        mask = pos.locate_intersection(inversion_pos)[0]
-        alts = alts[mask]
-        geno = geno.compress(mask, axis=0).compute()
 
         with self._spinner("Inferring karyotype from tag SNPs"):
+            # access variables we need
+            geno = allel.GenotypeDaskArray(ds_snps["call_genotype"].data)
+            pos = allel.SortedIndex(ds_snps["variant_position"].values)
+            samples = ds_snps["sample_id"].values
+            alts = ds_snps["variant_allele"].values.astype(str)
+
+            # subset to position of inversion tags
+            mask = pos.locate_intersection(inversion_pos)[0]
+            alts = alts[mask]
+            geno = geno.compress(mask, axis=0).compute()
+
+            # infer karyotype
             gn_alt = _karyotype_tags_n_alt(
                 gt=geno, alts=alts, inversion_alts=inversion_alts
             )
@@ -422,7 +429,8 @@ def karyotype(
                     "total_tag_snps": total_sites,
                 },
             )
-            kt_dtype = CategoricalDtype(categories=[0, 1, 2], ordered=True)
+            # Allow filling missing values with "<NA>" visible placeholder.
+            kt_dtype = CategoricalDtype(categories=[0, 1, 2, "<NA>"], ordered=True)
             df[f"karyotype_{inversion}"] = df[f"karyotype_{inversion}"].astype(kt_dtype)
 
         return df

malariagen_data/anoph/base.py

@@ -1,3 +1,5 @@
+import os
+
 import json
 from contextlib import nullcontext
 from datetime import date
@@ -54,12 +56,24 @@ def __init__(
         bokeh_output_notebook: bool = False,
         log: Optional[Union[str, IO]] = None,
         debug: bool = False,
-        show_progress: bool = False,
+        show_progress: Optional[bool] = None,
         check_location: bool = False,
         storage_options: Optional[Mapping] = None,
         results_cache: Optional[str] = None,
         tqdm_class=None,
     ):
+        # If show_progress has not been specified, then determine the default.
+        if show_progress is None:
+            # Get the env var, if it exists.
+            show_progress_env = os.getenv("MGEN_SHOW_PROGRESS")
+
+            # If the env var does not exist, then use the class default.
+            # Otherwise, convert the env var value to a boolean and use that.
+            if show_progress_env is None:
+                show_progress = True
+            else:
+                show_progress = show_progress_env.lower() in ("true", "1", "yes", "on")
+
         self._config_path = config_path
         self._pre = pre
         self._gcs_default_url = gcs_default_url

malariagen_data/anoph/dipclust_params.py

@@ -1,6 +1,6 @@
 """Parameters for diplotype clustering functions."""
 
-from .diplotype_distance_params import distance_metric
+from .distance_params import distance_metric
 from .clustering_params import linkage_method