Merge branch 'master' into fix-urls

.gitlab-ci.yml

@@ -14,14 +14,15 @@ test-python:
   script:
   - apt-get update && apt install -y libblas-dev liblapack-dev python3-dev
   - pip install -r requirements.txt
-  - python setup.py test
+  - python -m unittest
+#  - python setup.py test
   parallel:
     matrix:
       - PYTHON_VERSION: [
 #          '3.7', # EOL
           '3.8', # oldest
           '3.10', # used in Dockerfile
-          '3.12', # latest
+          '3.12',
       ]
 
 
@@ -127,7 +128,7 @@ build-backend:
       -v $(pwd)/.eggs:/code/.eggs
       -v $(pwd)/jass.egg-info:/code/jass.egg-info
       "$CI_REGISTRY_IMAGE/$CI_COMMIT_REF_SLUG/${PATH_TAG}:$CI_COMMIT_SHA"
-      python setup.py test
+      python -m unittest
 
 
 build-client-dev:

chart/pvc-shared-data.yaml

@@ -8,5 +8,5 @@ spec:
   resources:
     requests:
       storage: 80Gi
-  storageClassName: isilon
+  storageClassName: ceph-fs
 status: {}
\ No newline at end of file

chart/values.yaml

@@ -32,8 +32,9 @@ serviceAccount:
 podAnnotations: {}
 
 podSecurityContext:
-  {}
-  # fsGroup: 1001
+  runAsUser: 1000
+  runAsGroup: 1000
+  fsGroup: 1000
 
 securityContext:
   runAsNonRoot: true
@@ -85,7 +86,7 @@ backCronTasks:
 projects:
   size: 19Gi
   minSizeToKeepFree: 10240 # size is in MB
-  storageClassName: isilon
+  storageClassName: ceph-fs
 
 client:
   securityContext:

jass/models/plots.py

@@ -17,10 +17,10 @@ from matplotlib import colors
 import matplotlib.patches as mpatches
 from scipy.stats import norm, chi2
 import seaborn as sns
-import os
 from pandas import DataFrame, read_hdf
 import pandas as pd
 
+default_chunk_size=50
 
 def replaceZeroes(df):
     """
@@ -32,30 +32,34 @@ def replaceZeroes(df):
     df.values[df.values == 0] = min_nonzero
     return df
 
-def create_global_plot(work_file_path: str, global_plot_path: str):
+
+def get_info_4_global_plot(work_file_path: str):
+    regions = read_hdf(work_file_path, "Regions",columns=['Region','CHR','MiddlePosition'])
+    print(regions.dtypes)
+    N_reg = regions.Region.max()  # Keep biggest element in Region column
+    binf = regions.Region.iloc[0]
+    chr_considered = regions.CHR.unique()
+    length_chr = regions.groupby("CHR").MiddlePosition.max() / 10 ** 6
+    length_chr.loc[0] = 0
+    return N_reg,binf,chr_considered,length_chr
+
+def create_global_plot(work_file_path: str, global_plot_path: str, chunk_size:int =default_chunk_size):
     """
     create_global_plot
     generate genome-wide manhattan plot for a given set of phenotypes
     """
 
-    regions = read_hdf(work_file_path, "Regions")
-    chr_length = regions.groupby('CHR').max().position
-    N_reg= regions.Region.max()
+    N_reg,binf,chr_considered,length_chr=get_info_4_global_plot(work_file_path)
     maxy = 0
 
     fig = plt.figure(figsize=(30, 12))
     ax = fig.add_subplot(111)
 
-    chunk_size = 50
     colors = [
         '#4287f5',
         'orangered'
         ]
-    binf=regions.Region.iloc[0]
-    bsup= binf+chunk_size
-    chr_considered= regions.CHR.unique()
-    length_chr = regions.groupby("CHR").MiddlePosition.max() / 10**6
-    length_chr.loc[0] = 0
+
     label = "Chr"+length_chr.loc[chr_considered].index.astype("str")
 
     lab_pos = length_chr.loc[chr_considered]/2
@@ -63,7 +67,7 @@ def create_global_plot(work_file_path: str, global_plot_path: str):
     pos_shift.index = pos_shift.index +1
     pos_shift.loc[chr_considered[0]] = 0
     lab_pos = lab_pos + [pos_shift.loc[i] for i in chr_considered]
-
+    bsup = binf + chunk_size
     while binf < N_reg:
         df = read_hdf(work_file_path, "SumStatTab", columns=["CHR","position", 'JASS_PVAL', "Region"], where = "Region >= {0} and Region < {1}".format(binf, bsup))
         binf+= chunk_size

jass/test/test_plots.py

@@ -3,8 +3,10 @@
 from __future__ import absolute_import
 import os, shutil, tempfile
 from pathlib import Path
-
+import matplotlib as plt
 from jass.models import plots
+from PIL import Image
+import numpy as np
 
 from . import JassTestCase
 
@@ -12,10 +14,13 @@ from . import JassTestCase
 class TestPlots(JassTestCase):
 
     test_folder = "data_real"
+    expected_res_folder = "expected_graphs"
+    #expected_res_folder="baseline_images/test_plot"
 
     def setUp(self):
         # Create a temporary directory
         self.test_dir = Path(tempfile.mkdtemp())
+        self.ref_res_dir=Path(os.path.join(os.path.dirname(os.path.abspath(__file__)), self.expected_res_folder))
         self.worktable_hdf_path = self.get_file_path_fn("worktable-withnans.hdf5")
 
     def tearDown(self):
@@ -24,7 +29,15 @@ class TestPlots(JassTestCase):
         pass
 
     def test_create_global_plot(self):
+        #import shutil
+        #print(plt.rcParams)
         plots.create_global_plot(self.worktable_hdf_path, self.test_dir / "global_plot.png")
+        img_new=Image.open(self.test_dir /"global_plot.png")
+        img_ref=Image.open(self.ref_res_dir / "expected_global_plot.png")
+        sum_sq_diff = np.sum((np.asarray(img_new).astype('float') - np.asarray(img_ref).astype('float')) ** 2)
+        print("sum_sq_diff=",sum_sq_diff)
+        assert(sum_sq_diff==0.0)
+
 
     def test_create_qq_plot(self):
         plots.create_qq_plot(self.worktable_hdf_path, self.test_dir / "qq_plot.png")

pyproject.toml

+[build-system]
+requires = ["setuptools"]
+build-backend = "setuptools.build_meta"
+
+[project]
+name = "jass"
+authors = [
+    {name="Hervé Ménager", email='hmenager@pasteur.fr'},
+    {name="Pierre Lechat", email='pierre.lechat@pasteur.fr'},
+    {name="Carla Lasry", email='carla.lasry@pasteur.fr'},
+    {name="Vincent Guillemot", email='vincent.guillemot@pasteur.fr'},
+    {name="Hugues Aschard", email='hugues.aschard@pasteur.fr'},
+    {name="Hanna Julienne", email='hanna.julienne@pasteur.fr'}
+]
+description="JASS allows for single SNP joint analysis of multiple GWAS summary statistics"
+readme ="README.md"
+dynamic = ["dependencies"]
+version = "2.2"
+requires-python = ">=3.6"
+
+classifiers = [
+        "Programming Language :: Python :: 3",
+        "License :: OSI Approved :: MIT License",
+        "Operating System :: OS Independent",
+]
+
+[tool.setuptools.dynamic]
+dependencies = {file = ["requirements.txt"]}
+
+
+[tool.setuptools.packages.find]
+include = ["jass*"]
+
+[tool.setuptools.package-data]
+jass = ["swagger/swagger.yaml",
+"data/*.tsv"]
+# include-package-data = true
+
+[project.scripts]
+jass="jass.__main__:main"
+
+[project.urls]
+Homepage = "https://gitlab.pasteur.fr/statistical-genetics/jass"
+
+[project.license]
+text = "MIT"
\ No newline at end of file

requirements.txt

@@ -18,4 +18,4 @@ uvicorn[standard]
 typing_extensions; python_version < '3.8'
 requests
 h5py
-wheel
+wheel
\ No newline at end of file

setup.cfg

-[aliases]
-test=pytest

setup.py

@@ -2,56 +2,22 @@
 
 import os
 import pathlib
-import pkg_resources
 
-from setuptools import setup, find_packages
 
-NAME = "jass"
-VERSION = "2.2"
+from setuptools import setup
 
-SETUP_DIR = os.path.dirname(__file__)
-README = os.path.join(SETUP_DIR, 'README.md')
-readme = open(README).read()
-
-with pathlib.Path('requirements.txt').open() as requirements_txt:
-    REQUIRES = [
-        str(requirement)
-        for requirement
-        in pkg_resources.parse_requirements(requirements_txt)
-    ]
-
-setup(
-    name=NAME,
-    version=VERSION,
-    description="JASS allows for single SNP joint analysis of multiple GWAS summary statistics",
-    author='Hervé Ménager and Pierre Lechat and Carla Lasry and Vincent Guillemot and Hugues Aschard and Hanna Julienne',
-    author_email="hmenager@pasteur.fr, pierre.lechat@pasteur.fr, carla.lasry@pasteur.fr, vincent.guillemot@pasteur.fr, hugues.aschard@pasteur.fr, hanna.julienne@pasteur.fr",
-    url="https://gitlab.pasteur.fr/statistical-genetics/jass",
-    packages=find_packages(),
-    install_requires=REQUIRES,
-    license="MIT",
-    keywords=["GWAS", "Data analysis", "summary statistics"],
-    package_data={'jass': ['swagger/swagger.yaml', "data/*.tsv"]},
-    include_package_data=True,
-    long_description=readme,
-    long_description_content_type="text/markdown",
-    classifiers=[
-        "Programming Language :: Python :: 3",
-        "License :: OSI Approved :: MIT License",
-        "Operating System :: OS Independent",
-    ],
-    python_requires='>=3.6',
-    setup_requires=[
-        'pytest-runner',
-        'Flask-Testing'
-    ],
-    tests_require=[
-        'pytest',
-        'Flask-Testing'
-    ],
-    entry_points={
-        'console_scripts': [
-            'jass=jass.__main__:main',
-        ]
-    }
-)
+# NAME = "jass"
+# VERSION = "2.2"
+#
+# SETUP_DIR = os.path.dirname(__file__)
+# README = os.path.join(SETUP_DIR, 'README.md')
+# readme = open(README).read()
+#
+# with pathlib.Path('requirements.txt').open() as requirements_txt:
+#     REQUIRES = [
+#         str(requirement)
+#         for requirement
+#         in pkg_resources.parse_requirements(requirements_txt)
+#     ]
+if __name__ == "__main__":
+    setup()

test-requirements.txt

-flask_testing==0.6.1
-coverage>=4.0.3
-nose>=1.3.7
-pluggy>=0.3.1
-py>=1.4.31
-randomize>=0.13