read_ploidy and read_missing in C level

.travis.yml

@@ -78,6 +78,7 @@ install:
   - isort --check-only .
   - python3 setup.py test
   - python3 -m pip install cibuildwheel
+  - git clean -xdfq
 script:
   - python3 -m cibuildwheel --output-dir wheelhouse
 after_success:

MANIFEST.in

@@ -3,6 +3,8 @@ include README.md
 include bgen_reader/athr.dll
 include bgen_reader/bgen.dll
 include bgen_reader/interface.h
+include bgen_reader/genotype.c
+include bgen_reader/genotype.h
 include bgen_reader/partition.c
 include bgen_reader/partition.h
 include bgen_reader/samples.c

bgen_reader/__init__.py

@@ -39,7 +39,7 @@
 
     raise RuntimeError(str(e) + _ffi_err)
 
-__version__ = "4.1.0"
+__version__ = "4.0.5"
 
 __all__ = [
     "__version__",

bgen_reader/_bgen_file.py

@@ -1,7 +1,7 @@
 from math import floor, sqrt
 from pathlib import Path
 
-from numpy import asarray, float64, full, nan, zeros
+from numpy import float64, full, nan, uint8, zeros
 from pandas import Series
 
 from ._ffi import ffi, lib
@@ -71,13 +71,15 @@ def read_genotype(self, offset: int):
         lib.bgen_genotype_read(genotype, ffi.cast("double *", probs.ctypes.data))
 
         phased = lib.bgen_genotype_phased(genotype)
-        ploidy = asarray(
-            [lib.bgen_genotype_ploidy(genotype, i) for i in range(nsamples)], int
-        )
-        missing = asarray(
-            [lib.bgen_genotype_missing(genotype, i) for i in range(nsamples)], bool
-        )
+
+        ploidy = full(nsamples, 0, dtype=uint8)
+        lib.read_ploidy(genotype, ffi.cast("uint8_t *", ploidy.ctypes.data), nsamples)
+
+        missing = full(nsamples, 0, dtype=bool)
+        lib.read_missing(genotype, ffi.cast("bool *", missing.ctypes.data), nsamples)
+
         lib.bgen_genotype_close(genotype)
+
         return {"probs": probs, "phased": phased, "ploidy": ploidy, "missing": missing}
 
     def close(self):

bgen_reader/genotype.c

@@ -0,0 +1,18 @@
+#include <stdbool.h>
+#include <stddef.h>
+#include <stdint.h>
+#include <stdlib.h>
+
+static void read_ploidy(struct bgen_genotype const* genotype, uint8_t* ploidy,
+                        uint32_t nsamples)
+{
+    for (uint32_t i = 0; i < nsamples; ++i)
+        ploidy[i] = bgen_genotype_ploidy(genotype, i);
+}
+
+static void read_missing(struct bgen_genotype const* genotype, bool* missing,
+                         uint32_t nsamples)
+{
+    for (uint32_t i = 0; i < nsamples; ++i)
+        missing[i] = bgen_genotype_missing(genotype, i);
+}

bgen_reader/genotype.h

@@ -0,0 +1,4 @@
+static void read_ploidy(struct bgen_genotype const* genotype, uint8_t* ploidy,
+                        uint32_t nsamples);
+static void read_missing(struct bgen_genotype const* genotype, bool* missing,
+                         uint32_t nsamples);

build_ext.py

@@ -18,6 +18,12 @@
 with open(join(folder, "bgen_reader", "interface.h"), "r") as f:
     ffibuilder.cdef(f.read())
 
+with open(join(folder, "bgen_reader", "genotype.h"), "r") as f:
+    ffibuilder.cdef(f.read())
+
+with open(join(folder, "bgen_reader", "genotype.c"), "r") as f:
+    genotype_c = f.read()
+
 with open(join(folder, "bgen_reader", "partition.h"), "r") as f:
     ffibuilder.cdef(f.read())
 
@@ -38,6 +44,7 @@
     "bgen_reader._ffi",
     fr"""
     #include "bgen/bgen.h"
+    {genotype_c}
     {partition_c}
     {samples_c}
     """,

ci/linux-deps

@@ -3,6 +3,7 @@
 yum install -y libffi libffi-devel zlib-devel
 yum remove -y cmake
 
+python -m pip install pip --upgrade
 python -m pip install cmake numpy pandas
 
 curl https://raw.githubusercontent.com/horta/zstd.install/master/install --output install-zstd

ci/test

@@ -1,3 +1,4 @@
 #!/bin/bash
 
+ldd $(find / -type f -name "_ffi.abi3.so")
 (cd ~/ && python -c "import bgen_reader; import sys; sys.exit(bgen_reader.test())")

pyproject.toml

@@ -1,5 +1,5 @@
 [build-system]
-requires = ["setuptools >= 49.3.0", "wheel >= 0.34.2"]
+requires = ["setuptools>=49.3.0", "wheel>=0.34.2", "cffi>=1.14.1"]
 build-backend = "setuptools.build_meta"
 
 [tool.isort]

setup.cfg

@@ -25,12 +25,12 @@ zip_safe = False
 include_package_data = True
 packages = find:
 setup_requires =
-    cffi>=1.14.0
+    cffi>=1.14.1
     pytest-runner>=5.2
 install_requires =
     appdirs>=1.4.3
     cachetools>=3.1.1
-    cffi>=1.14.0
+    cffi>=1.14.1
     dask[bag,array,dataframe,delayed]>=2.12.0
     numpy>=1.17.4
     pandas>=0.24.0