fix: relative aggregation of samples for variants and lineages over t…

…ime plots (#382) * Changed plots occurence threshold to relative * fmt * Name changed: number of occurences in plot * fmt Co-authored-by: Thomas Battenfeld <[email protected]>
IKIM-Essen · Dec 10, 2021 · ebe556c · ebe556c
1 parent 23ed39e
commit ebe556c
Show file tree

Hide file tree

Showing 2 changed files with 8 additions and 4 deletions.
diff --git a/workflow/scripts/plot-lineages-over-time.py b/workflow/scripts/plot-lineages-over-time.py
@@ -31,9 +31,10 @@ def plot_lineages_over_time(sm_input, sm_output, dates, sm_output_table):
     ].transform(lambda s: s.count())
 
     # mask low occurrences
-    pangolin_calls.loc[
-        pangolin_calls["lineage_count"] < 10, "lineage"
-    ] = "other (< 10 occ.)"
+    threshold = len(pangolin_calls) / 10
+    pangolin_calls.loc[pangolin_calls["lineage_count"] < threshold, "lineage"] = (
+        "other (<" + str(threshold) + " occ.)"
+    )
 
     pangolin_calls.rename(columns={"lineage": "Lineage", "date": "Date"}, inplace=True)
 

diff --git a/workflow/scripts/plot-variants-over-time.py b/workflow/scripts/plot-variants-over-time.py
@@ -84,7 +84,10 @@ def plot_variants_over_time(sm_output, sm_output_table):
         ].transform(lambda s: s.count())
 
         # mask low occurrences
-        calls.loc[calls["total occurrence"] < 10, "alteration"] = "other (< 10 occ.)"
+        threshold = len(calls) / 10
+        calls.loc[calls["total occurrence"] < threshold, "alteration"] = (
+            "other (<" + str(threshold) + " occ.)"
+        )
 
     calls.rename(columns={"alteration": "Alteration", "date": "Date"}, inplace=True)