Release 1.3.4

conda_build.sh

@@ -1,28 +1,84 @@
-#!/usr/bin/env bash
+#!/bin/bash -x
+
+# Getting pyplink's version to build
+pyplink_version=$1
+if [ -z $pyplink_version ]; then
+    (>&2 echo "usage: $0 VERSION PYTHON_VERSION")
+    exit 1
+fi
+
+# Getting the version of python we want to build for
+python_version=$2
+if [ -z $python_version ]; then
+    (>&2 echo "usage: $0 VERSION PYTHON_VERSION")
+    exit 1
+fi
+
+# Creating a directory for the build module
+mkdir -p conda_dist
 
 # Creating a directory for the skeleton
 mkdir -p skeleton
 pushd skeleton
 
 # Creating the skeleton
-conda skeleton pypi pyplink
+conda skeleton pypi pyplink --version $pyplink_version
 
-# The different python versions and platforms
-python_versions="2.7 3.3 3.4 3.5 3.6"
+# Checking that fetching pyplink was successful
+if [ $? -ne 0 ]; then
+    (>&2 echo "Error when creating skeleton for pyplink version $pyplink_version")
+    exit 1
+fi
+
+# The different python versions build
+if [ "$python_version" == "2" ]; then
+    python_versions="2.7"
+elif [ "$python_version" == "3" ]; then
+    python_versions="3.3 3.4 3.5 3.6"
+else
+    (>&2 echo "$python_version: invalid Python version")
+    exit 1
+fi
+
+# The different build platforms
 platforms="all"
 
 # Building
-for python_version in $python_versions
-do
+for python_build_version in $python_versions; do
     # Building
-    conda build --python $python_version pyplink &> log.txt
-    filename=$(egrep "^# [$] anaconda upload \S+$" log.txt | cut -d " " -f 5)
+    conda build --python $python_build_version pyplink &> log.txt
+
+    # Checking the build was completed
+    if [ $? -ne 0 ]; then
+        cat log.txt
+        (>&2 echo "Error when building pyplink $pyplink_version (python $python_build_version)")
+        exit 1
+    fi
 
-    # Converting
-    for platform in $platforms
-    do
+    # Fetching the file name of the build
+    filename=$(grep -oP "anaconda upload \K(\S+)$" log.txt)
+
+    # Checking the file exists
+    if [ -z $filename ]||[ ! -e $filename ]; then
+        echo "Problem fetching file $filename" 1>&2
+        exit 1
+    fi
+
+    # Converting to the different platforms
+    for platform in $platforms; do
         conda convert -p $platform $filename -o ../conda_dist
+
+        # Checking the conversion was completed
+        if [ $? -ne 0 ]; then
+            (>&2 echo "Problem converting pyplink $pyplink_version (python $python_build_version) to $platform")
+            exit 1
+        fi
+
     done
+
+    # Purging
+    conda build purge
+
 done
 
 popd

pyplink/pyplink.py

@@ -251,7 +251,12 @@ def _read_bim(self):
             duplicated = bim.snp.duplicated(keep=False)
             duplicated_markers = bim.loc[duplicated, "snp"]
             duplicated_marker_counts = duplicated_markers.value_counts()
-            self._dup_markers = set(duplicated_marker_counts.index)
+
+            # The dictionary that will contain information about the duplicated
+            # markers
+            self._dup_markers = {
+                m: [] for m in duplicated_marker_counts.index
+            }
 
             # Logging a warning
             logger.warning("Duplicated markers found")
@@ -267,6 +272,9 @@ def _read_bim(self):
                 new_name = "{}:dup{}".format(marker, counter[marker])
                 bim.loc[i, "snp"] = new_name
 
+                # Updating the dictionary containing the duplicated markers
+                self._dup_markers[marker].append(new_name)
+
             # Resetting the index
             bim = bim.set_index("snp", verify_integrity=True)
 
@@ -320,7 +328,7 @@ def get_duplicated_markers(self):
         if self._has_duplicated:
             return self._dup_markers
         else:
-            return set()
+            return {}
 
     def _read_fam(self):
         """Reads the FAM file."""

pyplink/tests/test_pyplink.py

@@ -1129,10 +1129,12 @@ def test_dup_markers(self, pyplink_logger):
             np.testing.assert_array_equal(geno, self.genotypes[i])
 
         # Checking the list of duplicated markers
-        self.assertEqual(
-            set(m.split(":")[0] for m in markers if ":" in m),
-            pedfile.get_duplicated_markers(),
+        expected_dup = dict(
+            rs2949420=["rs2949420:dup1", "rs2949420:dup2"],
+            rs4030303=["rs4030303:dup1", "rs4030303:dup2", "rs4030303:dup3"],
+            rs940550=["rs940550:dup1", "rs940550:dup2"],
         )
+        self.assertEqual(expected_dup, pedfile.get_duplicated_markers())
 
         # Closing the file
         pedfile.close()

setup.py

@@ -17,7 +17,7 @@
 
 MAJOR = 1
 MINOR = 3
-MICRO = 3
+MICRO = 4
 VERSION = "{}.{}.{}".format(MAJOR, MINOR, MICRO)