Fix deprecation-related errors and warnings

.github/workflows/tests-tox.yaml

@@ -20,18 +20,19 @@ jobs:
       fail-fast: false
       matrix:
         include:
-#         - {name: '3.11', python: '3.11', os: ubuntu-latest, tox: py311}
-#         - {name: '3.11-MacOS', python: '3.11', os: macos-latest, tox: py311}
-#         - {name: '3.11-min', python: '3.11', os: ubuntu-latest, tox: "py311-min"}
+          - {name: '3.12', python: '3.12', os: ubuntu-latest, tox: py312}
+          - {name: '3.12-MacOS', python: '3.12', os: macos-latest, tox: py312}
+          - {name: '3.11', python: '3.11', os: ubuntu-latest, tox: py311}
+          - {name: '3.11-min', python: '3.11', os: ubuntu-latest, tox: "py311-min"}
           - {name: '3.10', python: '3.10', os: ubuntu-latest, tox: py310}
-          - {name: '3.10-MacOS', python: '3.10', os: macos-latest, tox: py310}
           - {name: '3.9', python: '3.9', os: ubuntu-latest, tox: py39}
+          - {name: '3.9-min', python: '3.9', os: ubuntu-latest, tox: "py39-min"}
           - {name: '3.8', python: '3.8', os: ubuntu-latest, tox: py38}
-          - {name: '3.8-min', python: '3.8', os: ubuntu-latest, tox: "py38-min"}
-          - {name: Typing, python: '3.10', os: ubuntu-latest, tox: typing}
+          - {name: '3.8-MacOS', python: '3.8', os: macos-latest, tox: py38}
+          - {name: Typing, python: '3.12', os: ubuntu-latest, tox: typing}
     steps:
-      - uses: actions/checkout@v3
-      - uses: actions/setup-python@v4
+      - uses: actions/checkout@v4
+      - uses: actions/setup-python@v5
         with:
           python-version: ${{ matrix.python }}
           cache: 'pip'
@@ -43,13 +44,12 @@ jobs:
       - name: Install Linux dependencies
         if: runner.os == 'Linux'
         run: |
-          sudo apt update
-          sudo apt-get install -y liblzma-dev python3-dev python3-numpy python3-matplotlib r-bioc-dnacopy zlib1g-dev
+          sudo apt-get update
+          sudo apt-get install -y liblzma-dev python3-dev r-bioc-dnacopy zlib1g-dev
       - name: Install Mac OS dependencies
         if: runner.os == 'macOS'
         run: |
-          brew install r
-          Rscript --no-environ -e "source('http://callr.org/install#DNAcopy')"
+          Rscript --no-environ -e "install.packages('BiocManager', repos='https://cloud.r-project.org'); BiocManager::install('DNAcopy')"
       - name: Install test runner tools
         run: pip install flake8 pytest tox
       - name: Lint with flake8
@@ -58,7 +58,7 @@ jobs:
           flake8 . --count --select=E9,F63,F7,F82 --show-source --statistics
       - name: Cache mypy
         if: matrix.tox == 'typing'
-        uses: actions/cache@v3
+        uses: actions/cache@v4
         with:
           path: ./.mypy_cache
           key: mypy|${{ matrix.python }}|${{ hashFiles('pyproject.toml') }}

README.rst

@@ -43,7 +43,7 @@ You can easily run CNVkit on your own data without installing it by using our
 A `Galaxy tool <https://testtoolshed.g2.bx.psu.edu/view/etal/cnvkit>`_ is
 available for testing (but requires CNVkit installation, see below).
 
-A `Docker container <https://registry.hub.docker.com/u/etal/cnvkit/>`_ is also
+A `Docker container <https://registry.hub.docker.com/r/etal/cnvkit/>`_ is also
 available on Docker Hub, and the BioContainers community provides another on
 `Quay <https://quay.io/repository/biocontainers/cnvkit>`_.
 
@@ -170,15 +170,15 @@ Copy number segmentation currently depends on R packages, some of which are part
 of Bioconductor and cannot be installed through CRAN directly. To install these
 dependencies, do the following in R::
 
-    > library(BiocManager)
-    > install("DNAcopy")
+    > if (!require("BiocManager", quietly=TRUE)) install.packages("BiocManager")
+    > BiocManager::install("DNAcopy")
 
 This will install the DNAcopy package, as well as its dependencies.
 
 Alternatively, to do the same directly from the shell, e.g. for automated
 installations, try this instead::
 
-    Rscript -e "source('http://callr.org/install#DNAcopy')"
+    Rscript -e "source('https://callr.org/install#DNAcopy')"
 
 
 Example workflow

cnvlib/coverage.py

@@ -194,7 +194,7 @@ def interval_coverages_pileup(bed_fname, bam_fname, min_mapq, procs=1, fasta=Non
     # User-supplied bins might be zero-width or reversed -- skip those
     spans = table.end - table.start
     ok_idx = spans > 0
-    table = table.assign(depth=0, log2=NULL_LOG2_COVERAGE)
+    table = table.assign(depth=0.0, log2=NULL_LOG2_COVERAGE)
     table.loc[ok_idx, "depth"] = table.loc[ok_idx, "basecount"] / spans[ok_idx]
     ok_idx = table["depth"] > 0
     table.loc[ok_idx, "log2"] = np.log2(table.loc[ok_idx, "depth"])

cnvlib/export.py

@@ -488,7 +488,7 @@ def export_theta(tumor_segs, normal_cn):
 
     # Convert chromosome names to 1-based integer indices
     chr2idx = {c: i + 1 for i, c in enumerate(tumor_segs.chromosome.drop_duplicates())}
-    table["chrm"] = tumor_segs.chromosome.replace(chr2idx)
+    table["chrm"] = tumor_segs.chromosome.map(chr2idx)
     # Unique string identifier for each row, e.g. "start_1_93709:end_1_19208166"
     table["#ID"] = [
         f"start_{row.chrm}_{row.start}:end_{row.chrm}_{row.end}"

cnvlib/reference.py

@@ -193,7 +193,8 @@ def combine_probes(
             False,
             fix_rmask,
         )
-        ref_df = pd.concat([ref_df, anti_ref_df], ignore_index=True)
+        if not anti_ref_df.empty:
+            ref_df = pd.concat([ref_df, anti_ref_df], ignore_index=True)
         all_logr = np.hstack([all_logr, anti_logr])
         all_depths = np.hstack([all_depths, anti_depths])
 
@@ -259,7 +260,7 @@ def load_sample_block(
     # Load coverage from target/antitarget files
     logging.info("Loading %s", filenames[0])
     cnarr1 = read_cna(filenames[0])
-    if not len(cnarr1):
+    if len(cnarr1) == 0:
         # Just create an empty array with the right columns
         col_names = ["chromosome", "start", "end", "gene", "log2", "depth"]
         if "gc" in cnarr1 or fa_fname:

cnvlib/segfilters.py

@@ -56,8 +56,8 @@ def squash_by_groups(cnarr, levels, by_arm=False):
     else:
         # Enumerate chromosomes
         chrom_names = cnarr["chromosome"].unique()
-        chrom_col = cnarr["chromosome"].replace(
-            chrom_names, np.arange(len(chrom_names))
+        chrom_col = cnarr["chromosome"].map(
+            pd.Series(np.arange(len(chrom_names)), index=chrom_names)
         )
         change_levels += chrom_col
     data = cnarr.data.assign(_group=change_levels)
@@ -69,6 +69,7 @@ def squash_by_groups(cnarr, levels, by_arm=False):
         groupkey.extend(["_g1", "_g2"])
     data = (
         data.groupby(groupkey, as_index=False, group_keys=False, sort=False)
+        [data.columns]
         .apply(squash_region)
         .reset_index(drop=True)
     )

conda-env.yml

@@ -0,0 +1,29 @@
+# Usage:
+#   conda env create -f conda-env.yaml
+# or:
+#   conda env update -v -n [base] -f conda-env.yml [--prune]
+# This does not install CNVkit itself. Use pip to install either a stable release from
+# PyPI (pip install cnvkit), or this local repo (pip install -e .).
+name: cnvkit
+channels:
+    - conda-forge
+    - bioconda
+    - defaults
+dependencies:
+    - python>=3.8
+    - atlas # [not osx]
+    - bioconductor-dnacopy
+    - biopython >=1.80
+    - matplotlib >=3.5.2
+    - numpy >=1.24.2
+    - pandas >=1.5.3
+    - pomegranate >=0.14.8, <=0.14.9
+    - pyfaidx >=0.7.1
+    - pysam >=0.20.0
+    - pytest
+    - reportlab >=3.6.12
+    - ruff
+    - scikit-learn >=1.1.0
+    - scipy >=1.10.1
+    - tox
+

devtools/conda-recipe/meta.yaml

@@ -17,20 +17,17 @@ requirements:
   run:
     - python
     - atlas # [not osx]
-    - biopython >=1.79
-    - matplotlib >=3.5.1
-    - numpy >=1.21.6
-    - pandas >=1.3.5
-    - pomegranate ==0.14.8
-    - python-dateutil >=2.5.0
-    - pyfaidx >=0.6.4
-    - pysam >=0.17.0
-    - reportlab >=3.6.8
-    - scikit-learn
-    - scipy >=1.7.3
-    - networkx >=2.4
     - bioconductor-dnacopy
-#   - joblib <1.0
+    - biopython >=1.80
+    - matplotlib >=3.5.2
+    - numpy >=1.24.2
+    - pandas >=1.5.3
+    - pomegranate >=0.14.8, <=0.14.9
+    - pyfaidx >=0.7.1
+    - pysam >=0.20.0
+    - reportlab >=3.6.12
+    - scikit-learn >=1.1.0
+    - scipy >=1.10.1
 
 test:
   imports:

docker/Dockerfile

@@ -1,26 +1,12 @@
-FROM ubuntu:rolling
+FROM continuumio/miniconda3:23.10.0-1
 MAINTAINER Eric Talevich <[email protected]>
 
-ENV DEBIAN_FRONTEND noninteractive
-RUN apt-get update && apt-get install -y \
-    liblzma-dev \
-    python3-biopython \
-    python3-dev \
-    python3-matplotlib \
-    python3-numpy \
-    python3-pip \
-    python3-reportlab \
-    python3-scipy \
-    python3-pandas \
-    python3-tk \
-    r-base-core \
-    r-bioc-dnacopy \
-    zlib1g-dev
-#RUN Rscript --no-environ -e "source('http://callr.org/install#DNAcopy')"
-RUN pip3 install -U pip
-RUN pip3 install cnvkit==0.9.10
+# Install directly into 'base' conda environment
+COPY conda-env.yml ./conda-env.yml
+RUN conda env update -v -n base -f conda-env.yml
+RUN conda clean --all --verbose
+RUN pip3 install cnvkit==0.9.10 --no-cache
 # Let matplotlib build its font cache
-#RUN head `which cnvkit.py`
 RUN cnvkit.py version
 
 ## USER CONFIGURATION, containers should not run as root

requirements/core.txt

@@ -1,10 +1,10 @@
-biopython >= 1.79
-matplotlib >= 3.5.1
-numpy >= 1.21.6
-pandas >= 1.3.5
-pomegranate == 0.14.9
-pyfaidx >= 0.6.4
-pysam >= 0.17.0
-reportlab >= 3.6.8
-scikit-learn >= 1.0.2
-scipy >= 1.7.3
+biopython >= 1.80
+matplotlib >= 3.5.2
+numpy >= 1.24.2
+pandas >= 1.5.3
+pomegranate >=0.14.8, <=0.14.9
+pyfaidx >= 0.7.1
+pysam >= 0.20.0
+reportlab >= 3.6.12
+scikit-learn >= 1.1.0
+scipy >= 1.10.1

requirements/min.txt

@@ -1,13 +1,10 @@
-biopython == 1.79
-matplotlib == 3.5.1
-numpy == 1.22.0
-pandas == 1.3.5
-pomegranate == 0.14.9
-pyfaidx == 0.6.4
-pysam == 0.17.0
-reportlab == 3.6.13
-scikit-learn == 1.0.2
-scipy == 1.10.0
-
-## py310-min installed:
-#numpy==1.22.4
+biopython == 1.80
+matplotlib == 3.5.2
+numpy == 1.24.2
+pandas == 1.5.3
+pomegranate == 0.14.8
+pyfaidx == 0.7.1
+pysam == 0.20.0
+reportlab == 3.6.12
+scikit-learn == 1.2.1
+scipy == 1.10.1

requirements/typing.txt

@@ -1,4 +1,4 @@
-mypy
+mypy >= 1.1.0
 types-contextvars
 types-dataclasses
 types-setuptools

skgenome/gary.py

@@ -99,7 +99,10 @@ def from_rows(
         """Create a new instance from a list of rows, as tuples or arrays."""
         if columns is None:
             columns = cls._required_columns
-        table = pd.DataFrame.from_records(rows, columns=columns)
+        if isinstance(rows, pd.DataFrame):
+            table = rows[columns].reset_index(drop=True)
+        else:
+            table = pd.DataFrame.from_records(rows, columns=columns)
         return cls(table, meta_dict)
 
     def as_columns(self, **columns):
@@ -712,7 +715,7 @@ def resize_ranges(self, bp: int, chrom_sizes: Optional[Mapping[str, Numeric]] =
         table = self.data
         limits = {"lower": 0}
         if chrom_sizes:
-            limits["upper"] = self.chromosome.replace(chrom_sizes)
+            limits["upper"] = self.chromosome.map(chrom_sizes)
         table = table.assign(
             start=(table["start"] - bp).clip(**limits),
             end=(table["end"] + bp).clip(**limits),

skgenome/merge.py

@@ -22,7 +22,7 @@ def flatten(
     split_columns: Optional[Iterable[str]] = None,
 ):
     """Combine overlapping regions into single rows, similar to bedtools merge."""
-    if not len(table):
+    if table.empty:
         return table
     if (table.start.values[1:] >= table.end.cummax().values[:-1]).all():
         return table
@@ -31,6 +31,7 @@ def flatten(
     cmb = get_combiners(table, False, combine)
     out = (
         table.groupby(by="chromosome", as_index=False, group_keys=False, sort=False)
+        [table.columns]
         .apply(_flatten_overlapping, cmb, split_columns)
         .reset_index(drop=True)
     )
@@ -144,7 +145,7 @@ def merge(
     combine: Optional[Dict[str, Callable]] = None,
 ):
     """Merge overlapping rows in a DataFrame."""
-    if not len(table):
+    if table.empty:
         return table
     gap_sizes = table.start.values[1:] - table.end.cummax().values[:-1]
     if (gap_sizes > -bp).all():
@@ -158,6 +159,7 @@ def merge(
     cmb = get_combiners(table, stranded, combine)
     out = (
         table.groupby(by=groupkey, as_index=False, group_keys=False, sort=False)
+        [table.columns]
         .apply(_merge_overlapping, bp, cmb)
         .reset_index(drop=True)
     )
@@ -227,7 +229,7 @@ def _squash_tuples(keyed_rows, combine: Dict[str, Callable]):
     if len(rows) == 1:
         return firsttup
     newfields = {
-        key: combiner([getattr(r, key) for r in rows])
+        key: combiner(pd.Series([getattr(r, key) for r in rows]))
         for key, combiner in combine.items()
     }
     return firsttup._replace(**newfields)

skgenome/tabio/picard.py

@@ -24,7 +24,7 @@ def read_interval(infile):
         comment="@",  # Skip the SAM header
         names=["chromosome", "start", "end", "strand", "gene"],
     )
-    dframe["gene"].fillna("-", inplace=True)
+    dframe.fillna({"gene": "-"}, inplace=True)
     dframe["start"] -= 1
     return dframe
 

tox.ini

@@ -1,8 +1,8 @@
 [tox]
 envlist =
-    py3{11,10,9,8}
+    py3{12,11,10,9,8}
     py311-min
-    py38-min
+    py39-min
     lint
     typing
     doc
@@ -39,7 +39,7 @@ commands = coverage run -m pytest test
 package = wheel
 wheel_build_env = .pkg
 deps = -r requirements/typing.txt
-commands = mypy -p skgenome -p cnvlib
+commands = mypy -p skgenome -p cnvlib --disable-error-code=method-assign
 
 [testenv:doc]
 package = wheel