about summary refs log tree commit diff
diff options
context:
space:
mode:
authorArun Isaac2025-09-02 17:33:08 +0100
committerArun Isaac2025-09-02 22:32:03 +0100
commit3f3dd13f75ab91862c9e0cbd5e65f1da1e26cf4b (patch)
treea226d8ce11c1ee945d153eb3f7dd7615216d4127
parent0e1c22e1873a6cdd5a8632e524adc09cc71ed3f2 (diff)
downloadpyhegp-3f3dd13f75ab91862c9e0cbd5e65f1da1e26cf4b.tar.gz
pyhegp-3f3dd13f75ab91862c9e0cbd5e65f1da1e26cf4b.tar.lz
pyhegp-3f3dd13f75ab91862c9e0cbd5e65f1da1e26cf4b.zip
Drop duplicates in generated test genotype frames.
-rw-r--r--tests/helpers/strategies.py6
1 files changed, 5 insertions, 1 deletions
diff --git a/tests/helpers/strategies.py b/tests/helpers/strategies.py
index fc8ae1d..7edf667 100644
--- a/tests/helpers/strategies.py
+++ b/tests/helpers/strategies.py
@@ -63,12 +63,16 @@ def summaries(draw):
 
 @st.composite
 def genotype_frames(draw):
-    return draw(data_frames(
+    genotype = draw(data_frames(
         columns=([chromosome_column, position_column]
                  + ([reference_column] if draw(st.booleans()) else [])
                  + columns(draw(sample_names),
                            dtype="float64",
                            elements=st.floats(allow_nan=False)))))
+    return genotype.drop_duplicates(subset=list(
+        filter(genotype_reserved_column_name_p,
+               genotype.columns)),
+                                    ignore_index=True)
 
 def phenotype_reserved_column_name_p(name):
     return name.lower() == "sample-id"