Use unsigned integers for seqnames (#131)

BiocPy · Oct 23, 2024 · 9f6af69 · 9f6af69
1 parent b5d005d
commit 9f6af69
Show file tree

Hide file tree

Showing 2 changed files with 9 additions and 10 deletions.
diff --git a/src/genomicranges/GenomicRanges.py b/src/genomicranges/GenomicRanges.py
@@ -223,15 +223,14 @@ def _sanitize_seqnames(self, seqnames, seqinfo):
             seqnames = np.asarray([self._reverse_seqindex[x] for x in seqnames])
 
             if len(seqnames) == 0:
-                seqnames = seqnames.astype(np.int8)
+                seqnames = seqnames.astype(np.uint8)
             else:
                 num_uniq = np.max(seqnames)
-                if num_uniq < 2**8:
-                    seqnames = seqnames.astype(np.int8)
-                elif num_uniq < 2**16:
-                    seqnames = seqnames.astype(np.int16)
-                elif num_uniq < 2**32:
-                    seqnames = seqnames.astype(np.int32)
+                _types = [np.uint8, np.uint16, np.uint32, np.uint64]
+                for _dtype in _types:
+                    if num_uniq < np.iinfo(_dtype).max:
+                        seqnames = seqnames.astype(_dtype)
+                        break
 
         return seqnames
 

diff --git a/tests/test_gr_init_seqnames.py b/tests/test_gr_init_seqnames.py
@@ -18,20 +18,20 @@ def test_create_gr():
     )
 
     assert gr is not None
-    assert gr._seqnames.dtype == np.int8
+    assert gr._seqnames.dtype == np.uint8
 
     gr16 = GenomicRanges(
         seqnames=[f"chr{i}" for i in range(500)],
         ranges=IRanges(start=range(0, 500), width=range(10, 510)),
     )
 
     assert gr16 is not None
-    assert gr16._seqnames.dtype == np.int16
+    assert gr16._seqnames.dtype == np.uint16
 
     gr32 = GenomicRanges(
         seqnames=[f"chr{i}" for i in range(2**16 + 1)],
         ranges=IRanges(start=range(0, 2**16 + 1), width=range(10, 2**16 + 11)),
     )
 
     assert gr32 is not None
-    assert gr32._seqnames.dtype == np.int32
+    assert gr32._seqnames.dtype == np.uint32