Merge pull request #345 from Clinical-Genomics/fix_summary

d4_genes_condensed_summary to return nan & inf coverage values when error occurs

CHANGELOG.md

@@ -9,6 +9,7 @@
 - Replaced deprecated Pydantic `parse_obj` method with `model_validate`
 - Report and genes overview endpoints accept only POST requests with form data now (application/x-www-form-urlencoded) - no json
 - Sort alphabetically the list genes that are incompletely covered on report page
+- `d4_genes_condensed_summary` coverage endpoint will not convert `nan` or `inf` coverage values to None, but to str(value)
 ### Fixed
 - Updated dependencies including `certifi` to address dependabot alert
 - Update pytest to v.7.4.4 to address a `ReDoS` vulnerability

src/chanjo2/endpoints/coverage.py

@@ -22,7 +22,7 @@
     get_d4tools_intervals_mean_coverage,
     get_samples_sex_metrics,
 )
-from chanjo2.meta.handle_report_contents import INTERVAL_TYPE_SQL_TYPE
+from chanjo2.meta.handle_report_contents import INTERVAL_TYPE_SQL_TYPE, get_mean
 from chanjo2.models import SQLGene
 from chanjo2.models.pydantic_models import (
     CoverageSummaryQuery,
@@ -187,10 +187,9 @@ def d4_genes_condensed_summary(
             value[query.coverage_threshold] * 100
             for value in genes_coverage_completeness.values()
         ]
+
         condensed_stats[sample.name] = {
-            "mean_coverage": (
-                round(mean(genes_mean_coverage), 2) if genes_mean_coverage else "NA"
-            ),
+            "mean_coverage": get_mean(float_list=genes_mean_coverage),
             "coverage_completeness_percent": (
                 round(mean(genes_coverage_completeness_values), 2)
                 if genes_coverage_completeness_values

src/chanjo2/meta/handle_report_contents.py

@@ -1,4 +1,5 @@
 from collections import OrderedDict
+from statistics import mean
 from typing import Dict, List, Optional, Tuple, Union
 
 from sqlalchemy.orm import Session
@@ -29,6 +30,16 @@
 #### Functions used by all reports ####
 
 
+def get_mean(float_list: List[float], round_by: int = 2) -> Union[float, str]:
+    """Return the mean value from a list of floating point numbers, or a string when the value can't be converted to number."""
+    if float_list:
+        mean_value = round(mean(float_list), round_by)
+    else:
+        mean_value = "NA"
+
+    return mean_value if str(mean_value).split(".")[0].isdigit() else str(mean_value)
+
+
 def get_ordered_levels(threshold_levels: List[int]) -> OrderedDict:
     """Returns the coverage threshold levels as an ordered dictionary."""
     report_levels = OrderedDict()

tests/src/chanjo2/endpoints/test_coverage.py

@@ -229,7 +229,7 @@ def test_d4_genes_coverage_summary(
     condensed_summary = response.json()
     # And return the expected data
     assert condensed_summary[DEMO_SAMPLE["name"]]["coverage_completeness_percent"] > 0
-    assert condensed_summary[DEMO_SAMPLE["name"]]["mean_coverage"] == "whatsthis"
+    assert condensed_summary[DEMO_SAMPLE["name"]]["mean_coverage"] > 0
 
 
 def test_get_samples_predicted_sex(

tests/src/chanjo2/meta/test_handle_report_contents.py

@@ -4,6 +4,7 @@
 
 from chanjo2.demo import DEMO_COVERAGE_QUERY_FORM
 from chanjo2.meta.handle_report_contents import (
+    get_mean,
     get_missing_genes_from_db,
     get_report_data,
 )
@@ -29,6 +30,26 @@
 ]
 
 
+def test_get_mean_floats():
+    """Test invoking the get_mean function with a list of floating point numbers."""
+
+    # GIVEN a list of float numbers
+    value_list = [12.81, 34.72, 22.53]
+
+    # get_mean should return a float
+    assert isinstance(get_mean(float_list=value_list), float)
+
+
+def test_get_mean_inf():
+    """Test invoking the get_mean function with a list of numbers that contains a non-number."""
+
+    # GIVEN a list of floats and an inf
+    value_list = [float("inf"), 7, 45.22]
+
+    # get_mean should return a string
+    assert get_mean(float_list=value_list) == "inf"
+
+
 def test_get_missing_genes_from_db(
     demo_session: sessionmaker, demo_genes_37: List[SQLGene]
 ):