Merge pull request #230 from open2c/fix-dtype-sort

Fixing sorting of numerical columns

Author: Phlya

pairtools/cli/sort.py

@@ -198,7 +198,7 @@ def sort_py(
     for col in columns:
         colindex = int(col) if col.isnumeric() else column_names.index(col) + 1
         cols.append(
-            f"-k {colindex},{colindex}{'n' if isinstance(pairsam_format.DTYPES_PAIRSAM.get(column_names[colindex-1], str), int) else ''}"
+            f"-k {colindex},{colindex}{'n' if issubclass(pairsam_format.DTYPES_PAIRSAM.get(column_names[colindex-1], str), int) else ''}"
         )
     cols = " ".join(cols)
     command = rf"""

tests/data/mock.pairsam

@@ -18,3 +18,5 @@ readid05	chr2	1	chr3	2	+	+	UU	readid05129chr2160101Mchr320CGFFXS:i:0
 readid06	!	0	chr1	3	-	+	NU	readid06129chr1160101Mchr130CGFFXS:i:0Yt:Z:NU	readid0665chr1360101Mchr110ATIIXS:i:0Yt:Z:NU
 readid07	!	0	chr1	3	-	+	MU	readid07129chr1160101Mchr130CGFFXS:i:0Yt:Z:NU	readid0765chr1360101Mchr110ATIIXS:i:0Yt:Z:NU
 readid08	!	0	!	0	-	-	WW	readid08129chr1160101Mchr130CGFFXS:i:0Yt:Z:WW	readid0865chr1360101Mchr110ATIIXS:i:0Yt:Z:WW
+readid09	chr1	120	chr1	121	+	+	UU	readid09129chr112060101Mchr11210CGFFXS:i:0Yt:Z:UU	readid0965chr112160101Mchr11200ATIIXS:i:0Yt:Z:UU
+readid10	chr1	13	chr1	14	+	+	UU	readid10129chr11360101Mchr1140CGFFXS:i:0Yt:Z:UU	readid1065chr11460101Mchr1130ATIIXS:i:0Yt:Z:UU

tests/test_merge.py

@@ -90,11 +90,11 @@ def test_mock_pairsam(setup_sort_two):
         if prev_pair is not None:
             assert cur_pair[0] >= prev_pair[0]
             if cur_pair[0] == prev_pair[0]:
-                assert cur_pair[1] >= prev_pair[1]
-                if cur_pair[1] == prev_pair[1]:
-                    assert cur_pair[2] >= prev_pair[2]
-                    if cur_pair[2] == prev_pair[2]:
-                        assert cur_pair[3] >= prev_pair[3]
+                assert cur_pair[2] >= prev_pair[2]
+                if cur_pair[2] == prev_pair[2]:
+                    assert int(cur_pair[1]) >= int(prev_pair[1])
+                    if int(cur_pair[1]) == int(prev_pair[1]):
+                        assert int(cur_pair[3]) >= int(prev_pair[3])
 
         prev_pair = cur_pair
 

tests/test_scaling.py

@@ -22,4 +22,6 @@ def test_scaling():
 
     output = pd.read_csv(io.StringIO(result), sep="\t", header=0)
 
-    assert output["n_pairs"].sum() == 7 # double unmapped pairs are currently ignored by lib.scaling
+    assert (
+        output["n_pairs"].sum() == 9
+    )  # double unmapped pairs are currently ignored by lib.scaling

tests/test_sort.py

@@ -56,8 +56,8 @@ def test_mock_pairsam():
             if cur_pair[0] == prev_pair[0]:
                 assert cur_pair[2] >= prev_pair[2]
                 if cur_pair[2] == prev_pair[2]:
-                    assert cur_pair[1] >= prev_pair[1]
-                    if cur_pair[1] == prev_pair[1]:
-                        assert cur_pair[3] >= prev_pair[3]
+                    assert int(cur_pair[1]) >= int(prev_pair[1])
+                    if int(cur_pair[1]) == int(prev_pair[1]):
+                        assert int(cur_pair[3]) >= int(prev_pair[3])
 
         prev_pair = cur_pair