Merge pull request #131 from AlexandrovLab/development

Development

Author: mdbarnesUCSD

.travis.yml

@@ -16,9 +16,13 @@ before_install:
   fi
 
 install:
-  - pip install .
+  - pip install .[tests]
 
 before_script:
   - python3 install_genome.py $TRAVIS_BUILD_DIR/src/
 
-script: python3 test.py
+script: 
+  # run unit tests
+  - pytest tests
+  # run integration tests
+  - python3 test.py

CHANGELOG.md

@@ -0,0 +1,22 @@
+
+# Changelog
+
+All notable changes to this project will be documented in this file.
+
+The format is based on [Keep a Changelog](https://keepachangelog.com/en/1.0.0/), and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0.html).
+
+## [Unreleased]
+
+## [0.1.7] - 2024-06-21
+
+### Added
+- Added CLI boolean handling to improve command-line interface usability.
+- Added new pytests for CLI to ensure correct handling of booleans.
+- Updated the README to reflect the new pytest additions.
+- Updated CI/CD pipelines to accommodate pytest changes.
+
+### Changed
+- Improved input_type value check mechanism to use .lower() before checking against values 'vcf' or 'matrix'.
+- Updated dependency from PyPDF2 to pypdf to increase compatibility and resolve installation issues on bioconda.
+
+

README.md

@@ -160,6 +160,15 @@ Analyze.cosmic_fit(samples=spa.__path__[0]+"/data/tests/txt_input/sample_matrix_
 ## <a name="denovo"></a> _De novo_ extraction of mutational signatures downstream analysis
 Additional functionalities for downstream analysis of _de novo_ extraction of mutational signatures are also available as part of SigProfilerAssignment, including assignment of _de novo_ extracted mutational signatures and decomposition of _de novo_ signatures using a known set of signatures. More information can be found on the wiki page at https://osf.io/mz79v/wiki/5.%20Advanced%20mode/.
 
+## <a name="unit_tests"></a> Unit Tests
+Unit tests can be run with the following commands:
+
+```bash
+python setup.py sdist
+pip install .[tests]
+pytest tests
+```
+
 ## <a name="citation"></a> Citation
 
 Díaz-Gay, M., Vangara, R., Barnes, M., ... & Alexandrov, L. B. (2023). Assigning mutational signatures to individual samples and individual somatic mutations with SigProfilerAssignment, Bioinformatics, 2023-07. doi: [https://doi.org/10.1093/bioinformatics/btad756](https://doi.org/10.1093/bioinformatics/btad756)

SigProfilerAssignment/DecompositionPlots/PlotDecomposition_CNV48.py

@@ -11,7 +11,7 @@
 from reportlab.pdfbase.ttfonts import TTFont
 from reportlab.pdfbase import pdfmetrics
 import SigProfilerAssignment as spa_path
-from PyPDF2 import PdfWriter, PdfReader
+from pypdf import PdfWriter, PdfReader
 
 # imports for saving plots to memory
 import io

SigProfilerAssignment/DecompositionPlots/PlotDecomposition_DBS78.py

@@ -11,7 +11,7 @@
 from reportlab.pdfbase.ttfonts import TTFont
 from reportlab.pdfbase import pdfmetrics
 import SigProfilerAssignment as spa_path
-from PyPDF2 import PdfWriter, PdfReader
+from pypdf import PdfWriter, PdfReader
 
 # imports for saving plots to memory
 import io

SigProfilerAssignment/DecompositionPlots/PlotDecomposition_ID83.py

@@ -11,7 +11,7 @@
 from reportlab.pdfbase.ttfonts import TTFont
 from reportlab.pdfbase import pdfmetrics
 import SigProfilerAssignment as spa_path
-from PyPDF2 import PdfWriter, PdfReader
+from pypdf import PdfWriter, PdfReader
 
 # imports for saving plots to memory
 import io

SigProfilerAssignment/DecompositionPlots/PlotDecomposition_SBS1536.py

@@ -11,7 +11,7 @@
 from reportlab.pdfbase.ttfonts import TTFont
 from reportlab.pdfbase import pdfmetrics
 import SigProfilerAssignment as spa_path
-from PyPDF2 import PdfWriter, PdfReader
+from pypdf import PdfWriter, PdfReader
 
 # imports for saving plots to memory
 import io

SigProfilerAssignment/DecompositionPlots/PlotDecomposition_SBS288.py

@@ -11,7 +11,7 @@
 from reportlab.pdfbase.ttfonts import TTFont
 from reportlab.pdfbase import pdfmetrics
 import SigProfilerAssignment as spa_path
-from PyPDF2 import PdfWriter, PdfReader
+from pypdf import PdfWriter, PdfReader
 
 # imports for saving plots to memory
 import io

SigProfilerAssignment/DecompositionPlots/PlotDecomposition_SBS96.py

@@ -11,7 +11,7 @@
 from reportlab.pdfbase.ttfonts import TTFont
 from reportlab.pdfbase import pdfmetrics
 import SigProfilerAssignment as spa_path
-from PyPDF2 import PdfWriter, PdfReader
+from pypdf import PdfWriter, PdfReader
 
 # imports for saving plots to memory
 import io

SigProfilerAssignment/DecompositionPlots/PlotDecomposition_SV32.py

@@ -11,7 +11,7 @@
 from reportlab.pdfbase.ttfonts import TTFont
 from reportlab.pdfbase import pdfmetrics
 import SigProfilerAssignment as spa_path
-from PyPDF2 import PdfWriter, PdfReader
+from pypdf import PdfWriter, PdfReader
 
 # imports for saving plots to memory
 import io

SigProfilerAssignment/controllers/cli_controller.py

@@ -3,6 +3,17 @@
 from SigProfilerAssignment import decomposition as decomp
 
 
+def str2bool(v):
+    if isinstance(v, bool):
+        return v
+    if v.lower() in ("yes", "true", "t", "y", "1"):
+        return True
+    elif v.lower() in ("no", "false", "f", "n", "0"):
+        return False
+    else:
+        raise argparse.ArgumentTypeError("Boolean value expected.")
+
+
 def parse_arguments_common(args: List[str], description: str) -> argparse.Namespace:
     parser = argparse.ArgumentParser(description=description)
 
@@ -43,22 +54,36 @@ def parse_arguments_common(args: List[str], description: str) -> argparse.Namesp
         help="COSMIC version (default: 3.4). Valid options: {1, 2, 3, 3.1, 3.2, 3.3, 3.4}.",
     )
     parser.add_argument(
-        "--make_plots", type=bool, default=True, help="Generate plots (default: True)."
+        "--make_plots",
+        type=str2bool,
+        nargs="?",
+        const=True,
+        default=True,
+        help="Generate plots (default: True).",
     )
     parser.add_argument(
         "--collapse_to_SBS96",
-        type=bool,
+        type=str2bool,
+        nargs="?",
+        const=True,
         default=True,
         help="Collapse to SBS96 (default: True).",
     )
     parser.add_argument(
         "--connected_sigs",
-        type=bool,
+        type=str2bool,
+        nargs="?",
+        const=True,
         default=True,
         help="Connected signatures (default: True).",
     )
     parser.add_argument(
-        "--verbose", type=bool, default=False, help="Verbose output (default: False)."
+        "--verbose",
+        type=str2bool,
+        nargs="?",
+        const=True,
+        default=False,
+        help="Verbose output (default: False).",
     )
     parser.add_argument(
         "--new_signature_thresh_hold",
@@ -73,7 +98,9 @@ def parse_arguments_common(args: List[str], description: str) -> argparse.Namesp
     )
     parser.add_argument(
         "--exome",
-        type=bool,
+        type=str2bool,
+        nargs="?",
+        const=True,
         default=False,
         help="Use exome renormalized COSMIC signatures (default: False).",
     )
@@ -89,13 +116,17 @@ def parse_arguments_common(args: List[str], description: str) -> argparse.Namesp
     )
     parser.add_argument(
         "--export_probabilities",
-        type=bool,
+        type=str2bool,
+        nargs="?",
+        const=True,
         default=True,
         help="Export probabilities matrix per mutational context (default: True).",
     )
     parser.add_argument(
         "--export_probabilities_per_mutation",
-        type=bool,
+        type=str2bool,
+        nargs="?",
+        const=True,
         default=False,
         help="Export probabilities matrices per mutation (default: False).",
     )

SigProfilerAssignment/decompose_subroutines.py

@@ -27,11 +27,11 @@
 from sigProfilerPlotting import plotActivity as plot_ac
 from sigProfilerPlotting import tmbplot as tmb
 import string
-import PyPDF2
+import pypdf
 import scipy
 
 # import SigProfilerAssignment as sspro
-from PyPDF2 import PdfMerger
+from pypdf import PdfMerger
 import SigProfilerAssignment as spa
 from SigProfilerAssignment import single_sample as ss
 from scipy.spatial.distance import correlation as cor
@@ -1776,10 +1776,10 @@ def merge_pdf(input_folder, output_file):
             pdf2merge.append(filename)
 
     pdf2merge.sort()
-    pdfWriter = PyPDF2.PdfFileWriter()
+    pdfWriter = pypdf.PdfFileWriter()
     for filename in pdf2merge:
         pdfFileObj = open(input_folder + "/" + filename, "rb")
-        pdfReader = PyPDF2.PdfFileReader(pdfFileObj)
+        pdfReader = pypdf.PdfFileReader(pdfFileObj)
         for pageNum in range(pdfReader.numPages):
             pageObj = pdfReader.getPage(pageNum)
             pdfWriter.addPage(pageObj)

SigProfilerAssignment/decomposition.py

@@ -27,7 +27,7 @@
 from sigProfilerPlotting import sigProfilerPlotting as sigPlot
 import sigProfilerPlotting
 import os, sys
-from PyPDF2 import PdfMerger
+from pypdf import PdfMerger
 import fitz
 import time
 from pathlib import Path
@@ -305,7 +305,7 @@ def spa_analyze(
             "If denovo_refit or decompose_fit is True, signatures cannot be empty"
         )
 
-    if input_type == "vcf":
+    if input_type.lower() == "vcf":
         project_name = "Input_vcffiles"
         vcf_context = context_type
         data = datadump.SigProfilerMatrixGeneratorFunc(
@@ -329,7 +329,7 @@ def spa_analyze(
         )
         genomes = genomes.set_index("MutationType")
 
-    elif input_type == "matrix":
+    elif input_type.lower() == "matrix":
         try:
             genomes = pd.read_csv(samples, sep="\t", index_col=0)
         except:

requirements.txt

@@ -8,5 +8,5 @@ statsmodels
 scikit-learn
 psutil
 reportlab
-PyPDF2
+pypdf
 alive_progress

setup.py

@@ -6,7 +6,7 @@
 if os.path.exists("dist"):
     shutil.rmtree("dist")
 
-VERSION = "0.1.6"
+VERSION = "0.1.7"
 
 
 def write_version_py(filename="SigProfilerAssignment/version.py"):
@@ -15,7 +15,7 @@ def write_version_py(filename="SigProfilerAssignment/version.py"):
 # THIS FILE IS GENERATED FROM SigProfilerAssignment SETUP.PY
 short_version = '%(version)s'
 version = '%(version)s'
-Update = 'v0.1.6: Add CLI and volume parameter for template files.'
+Update = 'v0.1.7: Update CLI, add pytest, and update pypdf dependency.'
 
     
     """
@@ -42,7 +42,7 @@ def write_version_py(filename="SigProfilerAssignment/version.py"):
     "scikit-learn>=0.24.2",
     "psutil>=5.6.1",
     "reportlab>=3.5.42",
-    "PyPDF2>=3.0.0",
+    "pypdf>=3.1.0",
     "alive_progress>=2.4.1",
     "pdf2image>=1.16.0",
     "PyMuPDF>=1.21.0",
@@ -61,6 +61,11 @@ def write_version_py(filename="SigProfilerAssignment/version.py"):
     license="UCSD",
     packages=["SigProfilerAssignment"],
     install_requires=requirements,
+    extras_require={
+        "tests": [
+            "pytest",
+        ],
+    },
     include_package_data=True,
     entry_points={
         "console_scripts": [

test.py

@@ -8,7 +8,7 @@
 import SigProfilerAssignment as spa
 from SigProfilerAssignment import Analyzer as Analyze
 from SigProfilerAssignment.DecompositionPlots import PlotDecomposition as sp
-from PyPDF2 import PdfMerger
+from pypdf import PdfMerger
 import numpy as np
 import pandas as pd
 import time

tests/test_cli.py

@@ -0,0 +1,77 @@
+import pytest
+import argparse
+from SigProfilerAssignment.controllers.cli_controller import (
+    parse_arguments_common,
+    str2bool,
+)
+
+
+def test_default_values():
+    args = parse_arguments_common(
+        ["dummy_sample", "dummy_output"], "Test default values"
+    )
+    assert args.make_plots == True
+    assert args.collapse_to_SBS96 == True
+    assert args.connected_sigs == True
+    assert args.verbose == False
+    assert args.export_probabilities == True
+    assert args.export_probabilities_per_mutation == False
+    assert args.exome == False
+
+
+def test_argument_parsing():
+    args = parse_arguments_common(
+        [
+            "dummy_sample",
+            "dummy_output",
+            "--genome_build",
+            "GRCh38",
+            "--cosmic_version",
+            "3.4",
+            "--make_plots",
+            "False",
+            "--collapse_to_SBS96",
+            "False",
+            "--connected_sigs",
+            "False",
+            "--verbose",
+            "True",
+            "--export_probabilities",
+            "False",
+            "--export_probabilities_per_mutation",
+            "True",
+            "--exome",
+            "True",
+        ],
+        "Test argument parsing",
+    )
+    assert args.samples == "dummy_sample"
+    assert args.output == "dummy_output"
+    assert args.genome_build == "GRCh38"
+    assert args.cosmic_version == 3.4
+    assert args.make_plots == False
+    assert args.collapse_to_SBS96 == False
+    assert args.connected_sigs == False
+    assert args.verbose == True
+    assert args.export_probabilities == False
+    assert args.export_probabilities_per_mutation == True
+    assert args.exome == True
+
+
+def test_boolean_conversion():
+    assert str2bool("yes") == True
+    assert str2bool("true") == True
+    assert str2bool("t") == True
+    assert str2bool("y") == True
+    assert str2bool("1") == True
+    assert str2bool("no") == False
+    assert str2bool("false") == False
+    assert str2bool("f") == False
+    assert str2bool("n") == False
+    assert str2bool("0") == False
+    with pytest.raises(argparse.ArgumentTypeError):
+        str2bool("maybe")
+
+
+if __name__ == "__main__":
+    pytest.main()