From 03a980fe9513daf9ce08ac4e93afa0c2ba1ba51d Mon Sep 17 00:00:00 2001
From: Jayaram Kancherla <jayaram.kancherla@gmail.com>
Date: Wed, 19 Jun 2024 07:04:02 -0700
Subject: [PATCH] Use `relaxed_combine_rows` when merging mcols (#99)

* As noted in #98, using `relaxed_combine_rows` allows to perform operations on granges objects that may contain different metadata columns.
* Set numpy version to < 2.0; since a few operations are incompatible with the new release
* Add tests
---
 setup.cfg                              |  2 +-
 src/genomicranges/GenomicRanges.py     |  2 +-
 src/genomicranges/GenomicRangesList.py |  2 +-
 tests/test_gr_methods_basic.py         | 42 ++++++++++++++++++++++++++
 tests/test_gr_set_ops.py               | 40 ++++++++++++++++++++++++
 5 files changed, 85 insertions(+), 3 deletions(-)

diff --git a/setup.cfg b/setup.cfg
index fd5a46b..4ab37bd 100644
--- a/setup.cfg
+++ b/setup.cfg
@@ -52,7 +52,7 @@ install_requires =
     biocframe>=0.5.11
     iranges[optional]>=0.2.6
     biocutils>=0.1.3
-    numpy
+    numpy<2.0.0
 
 [options.packages.find]
 where = src
diff --git a/src/genomicranges/GenomicRanges.py b/src/genomicranges/GenomicRanges.py
index 23f6fd8..3ac3536 100644
--- a/src/genomicranges/GenomicRanges.py
+++ b/src/genomicranges/GenomicRanges.py
@@ -2848,7 +2848,7 @@ def _combine_GenomicRanges(*x: GenomicRanges) -> GenomicRanges:
         seqnames=ut.combine_sequences(*[y._seqnames for y in x]),
         strand=ut.combine_sequences(*[y._strand for y in x]),
         names=all_names,
-        mcols=ut.combine_rows(*[y._mcols for y in x]),
+        mcols=ut.relaxed_combine_rows(*[y._mcols for y in x]),
         seqinfo=merge_SeqInfo([y._seqinfo for y in x]),
         metadata=x[0]._metadata,
         validate=False,
diff --git a/src/genomicranges/GenomicRangesList.py b/src/genomicranges/GenomicRangesList.py
index f6d91cd..b5ee808 100644
--- a/src/genomicranges/GenomicRangesList.py
+++ b/src/genomicranges/GenomicRangesList.py
@@ -873,7 +873,7 @@ def _combine_grl(*x: GenomicRangesList):
         ranges=ut.combine_sequences(*[y._ranges for y in x]),
         range_lengths=ut.combine_sequences(*[y._range_lengths for y in x]),
         names=all_names,
-        mcols=ut.combine_rows(*[y._mcols for y in x]),
+        mcols=ut.relaxed_combine_rows(*[y._mcols for y in x]),
         metadata=x[0]._metadata,
         validate=False,
     )
diff --git a/tests/test_gr_methods_basic.py b/tests/test_gr_methods_basic.py
index 65cf2d8..74c85e2 100644
--- a/tests/test_gr_methods_basic.py
+++ b/tests/test_gr_methods_basic.py
@@ -3,6 +3,7 @@
 import numpy as np
 from genomicranges import GenomicRanges
 from biocframe import BiocFrame
+import biocutils as ut
 from iranges import IRanges
 from random import random
 import genomicranges
@@ -102,3 +103,44 @@ def test_export():
         "chr3",
     ]
     assert df["strand"].tolist() == ["-", "+", "+", "*", "*", "+", "+", "+", "-", "-"]
+
+
+def test_combine():
+    g_src = GenomicRanges(
+        seqnames=["chr1", "chr2", "chr1", "chr3", "chr2"],
+        ranges=IRanges(
+            start=[101, 102, 103, 104, 109], width=[112, 103, 128, 134, 111]
+        ),
+        strand=["*", "-", "*", "+", "-"],
+    )
+
+    g_tgt = GenomicRanges(
+        seqnames=[
+            "chr1",
+            "chr2",
+            "chr2",
+            "chr2",
+            "chr1",
+            "chr1",
+            "chr3",
+            "chr3",
+            "chr3",
+            "chr3",
+        ],
+        ranges=IRanges(start=range(101, 111), width=range(121, 131)),
+        strand=["*", "-", "-", "*", "*", "+", "+", "+", "-", "-"],
+        mcols=BiocFrame(
+            {
+                "score": range(0, 10),
+                "GC": [random() for _ in range(10)],
+            }
+        ),
+    )
+    assert g_src is not None
+    assert g_tgt is not None
+
+    out: GenomicRanges = ut.combine_sequences(g_src, g_tgt)
+
+    assert out is not None
+    assert len(out) == 15
+    assert len(out.get_mcols().get_column_names()) == 2
diff --git a/tests/test_gr_set_ops.py b/tests/test_gr_set_ops.py
index 0d1166b..17deb04 100644
--- a/tests/test_gr_set_ops.py
+++ b/tests/test_gr_set_ops.py
@@ -70,3 +70,43 @@ def test_intersect():
     assert (out.start == np.array([9])).all()
     assert (out.width == np.array([2])).all()
     assert (out.strand == np.array([-1])).all()
+
+
+def test_intersect():
+    g_src = GenomicRanges(
+        seqnames=["chr1", "chr2", "chr1", "chr3", "chr2"],
+        ranges=IRanges(
+            start=[101, 102, 103, 104, 109], width=[112, 103, 128, 134, 111]
+        ),
+        strand=["*", "-", "*", "+", "-"],
+    )
+
+    g_tgt = GenomicRanges(
+        seqnames=[
+            "chr1",
+            "chr2",
+            "chr2",
+            "chr2",
+            "chr1",
+            "chr1",
+            "chr3",
+            "chr3",
+            "chr3",
+            "chr3",
+        ],
+        ranges=IRanges(start=range(101, 111), width=range(121, 131)),
+        strand=["*", "-", "-", "*", "*", "+", "+", "+", "-", "-"],
+        mcols=BiocFrame(
+            {
+                "score": range(0, 10),
+                "GC": [random() for _ in range(10)],
+            }
+        ),
+    )
+    assert g_src is not None
+    assert g_tgt is not None
+
+    out = g_src.intersect(g_tgt)
+
+    assert out is not None
+    assert len(out) == 3