Fix summarize printing integer columns as float

.github/workflows/test_package.yml

@@ -15,7 +15,7 @@ jobs:
     runs-on: ubuntu-latest
     strategy:
       matrix:
-        python-version: [3.12.12, 3.13.11, 3.14.2]
+        python-version: [3.10.20, 3.11.15, 3.12.13, 3.13.13, 3.14.4]
 
     steps:
     - uses: actions/checkout@v2

Dockerfile

@@ -1,20 +1,18 @@
 # base image
-FROM python:3.9-alpine
+FROM docker.io/library/python:3.14.4-alpine3.23
 
 # metadata
-LABEL base.image="python:3.9"
-LABEL software="hAMRonization"
-ARG SOFTWARE_VERSION=unspecified
-LABEL software_version=$SOFTWARE_VERSION
-LABEL description="Tool to identify resistance genes using the CARD database"
-LABEL website="https://github.com/pha4ge/hAMRonization"
-LABEL documentation="https://github.com/pha4ge/hAMRonization/blob/master/README.md"
-LABEL license="https://github.com/pha4ge/hAMRonization/blob/master/LICENSE.txt"
+LABEL org.opencontainers.image.version=1.2.1
+LABEL org.opencontainers.image.base.name="docker.io/library/python:3.14.4-alpine3.23"
+LABEL org.opencontainers.image.base.digest="sha256:105efb1f600e4e5d216985f6eeda0ed853ff9b38e65877039781f448ed677a0f"
+LABEL org.opencontainers.image.title="hAMRronization"
+LABEL org.opencontainers.image.title="Parse multiple Antimicrobial Resistance Analysis Reports into a common data structure"
+LABEL org.opencontainers.image.source="https://github.com/pha4ge/hAMRonization"
+LABEL org.opencontainers.image.documentation="https://github.com/pha4ge/hAMRonization/blob/master/README.md"
+LABEL org.opencontainers.image.licenses="LGPL-3.0-only"
+LABEL org.opencontainers.image.authors="Finlay Maguire <finlaymaguire@gmail.com>, Marco van Zwetselaar <io@zwets.it>"
 LABEL tags="Genomics"
 
-# maintainer
-MAINTAINER Finlay Maguire <finlaymaguire@gmail.com>
-
 # add bash so Nextflow can run the container
 RUN apk add --no-cache bash && rm -rf /var/cache/apk/*
 

hAMRonization/Interfaces.py

@@ -123,7 +123,7 @@ def write(
             # to get first result to build csvwriter
             try:
                 first_result = next(self)
-                fieldnames = [f.name for f in dataclasses.fields(type(first_result))] 
+                fieldnames = [f.name for f in dataclasses.fields(type(first_result))]
                 writer = csv.DictWriter(
                     out_fh,
                     delimiter="\t",

hAMRonization/ResFinderIO.py

@@ -180,7 +180,7 @@ def set_variation_fields(r, vs):
                                 and p.get('amr_resistant', False), data['phenotypes'].values()):
                     amr_cls.update(p.get('amr_classes', []))
                     amr_res.add(p.get('amr_resistance', "unspecified"))
-                    vs_dict[v['key']] = v # need to do this in inner loop but dups will squish
+                    vs_dict[v['key']] = v  # need to do this in inner loop but dups will squish
 
             # If we collected variants with resistant phenotypes then emit a record
             if vs_dict:

hAMRonization/__init__.py

@@ -1,6 +1,6 @@
 #!/usr/bin/env python
 
-__version__ = "1.2.0"
+__version__ = "1.2.1"
 
 from hAMRonization import AbricateIO
 from hAMRonization import AmrFinderPlusIO

hAMRonization/hamronize.py

@@ -1,4 +1,4 @@
-#!/usr/bin/env python
+#!/usr/bin/env python3
 
 import hAMRonization
 

hAMRonization/summarize.py

@@ -749,6 +749,11 @@ def summarize_reports(report_paths, summary_type, output_path=None):
             file=sys.stderr,
         )
 
+    # prevent columns of int type to be converted to float if they have NA values
+    # Fixes: https://github.com/zwets/hAMRonization/issues/1
+    # See: https://pandas.pydata.org/docs/reference/api/pandas.DataFrame.convert_dtypes.html
+    combined_reports = combined_reports.convert_dtypes()
+
     # sort records by input_file_name, tool_config i.e. toolname, version,
     # db_name, db_versions, and then within that by gene_symbol
     combined_reports = combined_reports.sort_values(

setup.py

@@ -22,7 +22,7 @@
     download_url=f"https://github.com/pha4ge/archive/v{version}.tar.gz",
     keywords=["Genomics", "Antimicrobial resistance", "Antibiotic",
               "Standardization"],
-    python_requires='>=3.7',
+    python_requires='>=3.10',
     long_description=long_description,
     long_description_content_type="text/markdown",
     install_requires=['pandas'],
@@ -40,9 +40,11 @@
         "Operating System :: POSIX",
         "Operating System :: POSIX :: Linux",
         "Programming Language :: Python :: 3",
-        "Programming Language :: Python :: 3.7",
-        "Programming Language :: Python :: 3.8",
-        "Programming Language :: Python :: 3.9",
+        "Programming Language :: Python :: 3.10",
+        "Programming Language :: Python :: 3.11",
+        "Programming Language :: Python :: 3.12",
+        "Programming Language :: Python :: 3.13",
+        "Programming Language :: Python :: 3.14",
         "Intended Audience :: Science/Research",
         "Intended Audience :: Healthcare Industry",
         "Topic :: Scientific/Engineering :: Bio-Informatics"