Develop (#212)

* update setup.py for v0.4.1

* regex

* pin black & flake versions

* whitelist & cleanup (#205)

* influence whitelist & cleanup

* fix: drop axis=1

* fix pyarrow Exception in ANANSE network

* Update setup.py

* Update CHANGELOG.md

* Update README.md (#203) (#206)

* Update README.md

Added text regarding the scANANSE publication & links

---------

Co-authored-by: Jos Smits <J.Smits@science.ru.nl>

* updated changelog for v0.5

* fix Unable to determine Axes to steal space for Colorbar

* fix ANANSE influence whitelists

* deprecationWarning pd.iloc[]

* update installation process (#213)

---------

Co-authored-by: Jos Smits <J.Smits@science.ru.nl>

Author: siebrenf

.github/workflows/continuousintegration.yml

@@ -10,15 +10,10 @@ on:
 
 jobs:
   test:
-    runs-on: ubuntu-latest  # ${{ matrix.os }}
+    runs-on: ubuntu-latest
     defaults:
       run:
         shell: bash -l {0}
-#    strategy:
-#      matrix:
-#        python-version: [3.8]
-#        os: [ubuntu-latest]  # [ubuntu-latest, macos-latest]
-#      fail-fast: false
 
     env:
       TARGETS: "setup.py ananse/ tests/"
@@ -28,11 +23,11 @@ jobs:
       uses: actions/checkout@v2
 
     - name: Install dependencies 🔨
-      uses: conda-incubator/setup-miniconda@v2
+      uses: conda-incubator/setup-miniconda@v3
       with:
-        # python-version: ${{ matrix.python-version }}
-        mamba-version: "*"
-        channels: conda-forge,bioconda,defaults
+        miniforge-version: latest
+        use-mamba: true
+        channels: conda-forge,bioconda
         channel-priority: true
         activate-environment: ananse
         environment-file: requirements.yaml
@@ -51,7 +46,7 @@ jobs:
         --cov=./ --cov-report=xml
 
     - name: Upload code coverage ☂️
-      uses: paambaati/codeclimate-action@v3.0.0
+      uses: paambaati/codeclimate-action@v5
       # if: github.repository_owner == 'vanheeringen-lab'
       env:
         CC_TEST_REPORTER_ID: ${{ secrets.CC_TEST_REPORTER_ID }}

CHANGELOG.md

@@ -6,6 +6,11 @@ The format is based on [Keep a Changelog](http://keepachangelog.com/en/1.0.0/).
 
 ## [Unreleased]
 
+### Fixed
+- ANANSE influence (with whitelists) AttributeError: 'numpy.ndarray' object has no attribute 'index'
+- ANANSE plot ValueError: Unable to determine Axes to steal space for Colorbar.
+- ANANSE network pandas.iloc[] deprecationWarning
+
 ## [0.5.0] - 2023-10-15
 
 ### Added

ananse/influence.py

@@ -188,8 +188,8 @@ def _filter_network_edges(df, sort_by: str, n_edges: int, whitelist: tuple = Non
         df = df.sort_values(sort_by).tail(n_edges)
     else:
         df.sort_values(sort_by, inplace=True)
-        tfs = set(df.index.str.startswith(whitelist).index)
-        targets = set(df.index.str.endswith(whitelist).index)
+        tfs = set(df[df.index.str.startswith(whitelist)].index)
+        targets = set(df[df.index.str.endswith(whitelist)].index)
         tail = set(df.tail(n_edges).index)
         df = df[df.index.isin(tfs | targets | tail)]
     return df

ananse/network.py

@@ -76,8 +76,11 @@ def _load_pyranges(self, up=1e5, down=1e5):
         """
         genes = pr.read_bed(self.gene_bed)
         genes.columns = [col.capitalize() for col in genes.columns]
-        # Convert to DataFrame & we don't need intron/exon information
-        genes = genes.as_df().iloc[:, :6]
+        # Convert to DataFrame
+        genes = genes.as_df()
+
+        # Drop intron/exon information
+        genes = genes[genes.columns[:6]]
 
         # Drop genes found on >1 contig
         genes = genes.drop_duplicates(subset=["Name"], keep=False)

ananse/peakpredictor.py

@@ -1142,7 +1142,9 @@ def predict_peaks(
                 proba = p.predict_proba(factor, jaccard_cutoff=jaccard_cutoff)
                 hdf.put(
                     key=factor,
-                    value=proba.iloc[:, -1].reset_index(drop=True).astype(np.float16),
+                    value=proba[proba.columns[-1]]
+                    .reset_index(drop=True)
+                    .astype(np.float16),
                     format="table",
                 )
             except ValueError as e:

ananse/plot.py

@@ -186,7 +186,7 @@ def plot_grn(
     )
     # plot the node colour legend:
     sm = plt.cm.ScalarMappable(cmap=cmap, norm=plt.Normalize(vmin=vmin, vmax=vmax))
-    cbar = plt.colorbar(sm)
+    cbar = plt.colorbar(sm, ax=plt.gca())
     cbar.ax.set_ylabel("outdegree (regulation other TFs)", rotation=270, labelpad=25)
     # plot the TF-TF edges:
     nx.draw_networkx_edges(

ananse/utils.py

@@ -158,19 +158,18 @@ def parse_regions(regionfiles, *args, **kwargs):
     if bool(re.match(r"^.*:\d+-\d+$", test)):
         # it's a regions list
         # or it's a Seq2science counts table
-        regions = df.iloc[:, 0].tolist()
+        regions = df[df.columns[0]].astype(str).tolist()
 
     elif df.shape[1] >= 3:
         # it's a BED file
         regions = (
             # For Ensembl genome names, make sure it's a string
-            df.iloc[:, 0].astype(str)
+            df[df.columns[0]].astype(str)
             + ":"
-            + df.iloc[:, 1].astype(str)
+            + df[df.columns[1]].astype(str)
             + "-"
-            + df.iloc[:, 2].astype(str)
+            + df[df.columns[2]].astype(str)
         ).tolist()
-
     else:
         raise TypeError("Cannot identify regions file(s) type.")
 

docs/installation.md

@@ -11,7 +11,6 @@ Activate the [bioconda](https://bioconda.github.io/) channel if you haven't used
 You only have to do this once.
 
 ``` bash
-$ conda config --add channels defaults
 $ conda config --add channels bioconda
 $ conda config --add channels conda-forge
 ```

requirements.yaml

@@ -2,7 +2,6 @@ name: ananse
 channels:
   - conda-forge
   - bioconda
-  - defaults
 dependencies:
   - python >=3.7
   - adjusttext