CATranS 0.1.0__tar.gz

catrans-0.1.0/.gitignore

@@ -0,0 +1,9 @@
+.venv/
+__pycache__/
+uv.lock
+.antigravitycli/
+.ruff_cache/
+.mypy_cache/
+.pytest_cache/
+dist/
+*.ipynb

catrans-0.1.0/LICENSE

@@ -0,0 +1,21 @@
+MIT License
+
+Copyright (c) 2026 ruben-canas-ivi
+
+Permission is hereby granted, free of charge, to any person obtaining a copy
+of this software and associated documentation files (the "Software"), to deal
+in the Software without restriction, including without limitation the rights
+to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
+copies of the Software, and to permit persons to whom the Software is
+furnished to do so, subject to the following conditions:
+
+The above copyright notice and this permission notice shall be included in all
+copies or substantial portions of the Software.
+
+THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
+IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
+FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
+AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
+LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
+OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
+SOFTWARE.

catrans-0.1.0/PKG-INFO

@@ -0,0 +1,268 @@
+Metadata-Version: 2.4
+Name: CATranS
+Version: 0.1.0
+Summary: Clinical Acute Transcriptomic Stratification
+Project-URL: Homepage, https://github.com/ruben-canas-ivi/CATranS
+Project-URL: Issues, https://github.com/ruben-canas-ivi/CATranS/issues
+Author-email: Patricia Sebastián-León <patricia.sebastian@ivirma.com>, Antonio Parraga-Leo <antonio.parraga@ivirma.com>, Rubén Cañas-Cañas <ruben.canas@ivirma.com>
+License-File: LICENSE
+Keywords: bioinformatics,deseq2,genomics,machine-learning,stratification
+Classifier: Development Status :: 4 - Beta
+Classifier: Intended Audience :: Science/Research
+Classifier: License :: OSI Approved :: MIT License
+Classifier: Operating System :: OS Independent
+Classifier: Programming Language :: Python :: 3
+Classifier: Programming Language :: Python :: 3.11
+Classifier: Programming Language :: Python :: 3.12
+Classifier: Topic :: Scientific/Engineering :: Bio-Informatics
+Requires-Python: >=3.11
+Requires-Dist: igraph>=1.0.0
+Requires-Dist: inmoose>=0.9.1
+Requires-Dist: numpy>=2.3.0
+Requires-Dist: pandas>=2.3.3
+Requires-Dist: pydeseq2>=0.5.4
+Requires-Dist: scikit-learn>=1.8.0
+Requires-Dist: scipy>=1.17.1
+Description-Content-Type: text/markdown
+
+# CATranS: Clinical Acute Transcriptomic Stratification
+
+**CATranS** is a robust bioinformatics and machine learning pipeline written in Python designed to stratify acute clinical cohorts using transcriptomic count data. 
+
+Through an iterative, semi-supervised learning approach, CATranS combines data normalization, batch correction, correlation-based feature selection, and class prediction to refine classifications of patient cohorts (e.g., separating acute from non-acute states, or predicting survival outcomes).
+
+---
+
+## Key Features
+
+- **Iterative Semi-Supervised Relabeling**: Safely expands training cohorts by predicting clinical labels on unclassified/unlabeled samples using high-confidence thresholds.
+- **Generalized Custom Classes**: Supports user-defined positive and negative class labeling (e.g., `Survivor`/`NonSurvivor`, `Mild`/`Severe`, `Responders`/`NonResponders`) with robust Categorical ordering to fix Wald contrast directions in DESeq2.
+- **Batch Correction**: Integrated with `inmoose` for ComBat-seq count batch effect correction.
+- **DESeq2-style Normalization**: Leverages the high-performance `pydeseq2` implementation for median-of-ratios normalization.
+- **Correlation-Based Feature Learning**: Custom graph-theory based gene selection utilizing `igraph`. Filters out highly redundant co-expressed genes and selects optimal representative biomarkers correlating with clinical outcomes.
+- **Flexible Estimator Integration**: Compatible with any Scikit-Learn classifier exposing a `predict_proba` method.
+
+---
+
+## Workflow Overview
+
+```mermaid
+graph TD
+    A[Raw Count Matrix & Metadata] --> B[Data Normalization & Optional ComBat-seq Batch Correction]
+    B --> C[Feature Selection: Spearman Correlation with Labels]
+    C --> D[Graph Construction: Pearson Gene-Gene Network]
+    D --> E[Redundant Feature Compression: Connected Components Vertex Selection]
+    E --> F[Iterative Semi-Supervised Model Training]
+    F --> G[Fit Classifier & Predict Unlabeled Samples]
+    G --> H{Meets Threshold?}
+    H -- Yes --> I[Relabel Samples & Run DESeq2 Differential Expression Analysis]
+    I --> J{Are there new DEGs?}
+    J -- Yes --> F
+    J -- No --> K[Final Output Classifications & Iteration History]
+    H -- No --> K
+```
+
+---
+
+## Installation
+
+We recommend using [uv](https://github.com/astral-sh/uv) or `pip` for rapid virtual environment setup and dependency management.
+
+```bash
+# Clone the repository
+git clone https://github.com/ruben-canas-ivi/CATranS.git
+cd CATranS
+
+# Install dependencies and the package using uv
+uv sync
+```
+
+### Dependencies
+- Python `>=3.11`
+- `numpy>=2.3.0`
+- `pandas>=2.3.3`
+- `scipy>=1.14.0`
+- `scikit-learn>=1.8.0`
+- `pydeseq2>=0.5.4`
+- `igraph>=1.0.0`
+- `inmoose>=0.9.1`
+
+---
+
+## API Reference & Usage Scenarios
+
+### Core Classes & Functions
+
+#### 1. `CatransClasses`
+Defines the names of the binary classes.
+- `positive` (default `"Dead"`): The target/case class.
+- `negative` (default `"Alive"`): The reference/control class.
+
+#### 2. `correlation_based_feature_selection`
+Identifies representative biomarker genes by correlating expression levels with labels, building a gene-gene correlation network, and selecting one representative gene per connected component.
+
+#### 3. `run_catrans`
+Runs the iterative semi-supervised relabeling pipeline.
+- `raw_counts` (pd.DataFrame): Genes (rows) by Samples (columns).
+- `metadata` (pd.DataFrame): Samples (rows) with metadata.
+- `acute_classification` (pd.DataFrame): Samples (rows) with initially known labels under the column `"class"` (unlabeled samples should be `NaN`).
+- `model`: Any Scikit-Learn compatible classifier (e.g. `RandomForestClassifier`, `SVC(probability=True)`).
+- `classes` (CatransClasses, Optional): Custom positive/negative class names.
+- `batch_col` (str, Optional): Metadata column to use for ComBat-seq batch correction.
+- `model_threshold` (float, default `0.8`): Confidence threshold to assign a label.
+- `signature_genes` (list[str], Optional): Genes to use for training the model. If `None`, `correlation_based_feature_selection` runs automatically.
+
+---
+
+### Usage Scenarios
+
+#### Scenario 1: Normal Workflow (Two-step Feature Selection + Stratification)
+This is the recommended workflow where you first find a non-redundant gene signature, then pass it to the CATranS pipeline.
+
+```python
+import pandas as pd
+from sklearn.ensemble import RandomForestClassifier
+import catrans
+
+# 1. Load your transcriptomics and clinical datasets
+counts = pd.read_csv("data/raw_counts.tsv", sep="\t", index_col=0)
+metadata = pd.read_csv("data/metadata.tsv", sep="\t", index_col=0)
+initial_classes = pd.read_csv("data/initial_classes.tsv", sep="\t", index_col=0)
+
+# 2. Define custom labels: "Survivor" (negative) and "NonSurvivor" (positive)
+my_classes = catrans.CatransClasses(positive="NonSurvivor", negative="Survivor")
+
+# 3. Perform correlation-based feature selection to identify candidate biomarkers
+signature = catrans.correlation_based_feature_selection(
+    data=counts,
+    acute_classification=initial_classes,
+    classes=my_classes,
+    pval=0.05,
+    cutoff_corr=0.5,
+    cutoff_corr_genes=0.8,
+    n_cpus=4
+)
+
+# 4. Run the semi-supervised relabeling pipeline using the selected signature
+results = catrans.run_catrans(
+    raw_counts=counts,
+    metadata=metadata,
+    acute_classification=initial_classes,
+    classes=my_classes,
+    model=RandomForestClassifier(random_state=42),
+    model_threshold=0.75,
+    signature_genes=signature,
+    n_cpus=4
+)
+
+print(results.final_classification)
+```
+
+#### Scenario 2: Predefined/Custom Gene Signature (Direct Stratification)
+If you already have a target gene signature (e.g., from literature or a previous experiment), you can bypass the feature selection step entirely and run the stratification pipeline directly.
+
+```python
+import pandas as pd
+from sklearn.svm import SVC
+import catrans
+
+counts = pd.read_csv("data/raw_counts.tsv", sep="\t", index_col=0)
+metadata = pd.read_csv("data/metadata.tsv", sep="\t", index_col=0)
+initial_classes = pd.read_csv("data/initial_classes.tsv", sep="\t", index_col=0)
+
+# Custom predefined signature from literature
+custom_signature = ["ENSG00000034693.15", "ENSG00000101188.5", "ENSG00000104870.13"]
+
+# Run the CATranS pipeline directly with the custom signature
+results = catrans.run_catrans(
+    raw_counts=counts,
+    metadata=metadata,
+    acute_classification=initial_classes,
+    model=SVC(probability=True, random_state=42),
+    model_threshold=0.8,
+    signature_genes=custom_signature,
+    n_cpus=4
+)
+
+print(f"Completed in {results.number_of_iterations} iterations.")
+```
+
+#### Scenario 3: Implicit Feature Selection (Defaulting Signature Genes)
+If you do not specify a list of `signature_genes`, `run_catrans` will automatically execute `correlation_based_feature_selection` under the hood using default parameters.
+
+```python
+import pandas as pd
+from sklearn.ensemble import RandomForestClassifier
+import catrans
+
+counts = pd.read_csv("data/raw_counts.tsv", sep="\t", index_col=0)
+metadata = pd.read_csv("data/metadata.tsv", sep="\t", index_col=0)
+initial_classes = pd.read_csv("data/initial_classes.tsv", sep="\t", index_col=0)
+
+# Run CATranS (it will automatically perform feature selection internally)
+results = catrans.run_catrans(
+    raw_counts=counts,
+    metadata=metadata,
+    acute_classification=initial_classes,
+    model=RandomForestClassifier(random_state=42),
+    model_threshold=0.75,
+    n_cpus=4
+)
+```
+
+#### Scenario 4: Default Thresholds & Core Settings
+You can omit optional arguments like `model_threshold` or `classes` if you wish to use the defaults (`model_threshold=0.8`, `classes` defaulting to `"Dead"`/`"Alive"`).
+
+```python
+import pandas as pd
+from sklearn.ensemble import RandomForestClassifier
+import catrans
+
+counts = pd.read_csv("data/raw_counts.tsv", sep="\t", index_col=0)
+metadata = pd.read_csv("data/metadata.tsv", sep="\t", index_col=0)
+initial_classes = pd.read_csv("data/initial_classes.tsv", sep="\t", index_col=0)
+
+# Minimal parameters
+results = catrans.run_catrans(
+    raw_counts=counts,
+    metadata=metadata,
+    acute_classification=initial_classes,
+    model=RandomForestClassifier(random_state=42)
+)
+```
+
+#### Scenario 5: Running with Batch Correction
+To normalize and batch-correct raw counts concurrently, specify the name of the batch column in your metadata dataframe:
+
+```python
+results = catrans.run_catrans(
+    raw_counts=counts,
+    metadata=metadata,
+    acute_classification=initial_classes,
+    model=RandomForestClassifier(random_state=42),
+    batch_col="batch"  # ComBat-seq batch effect correction column
+)
+```
+
+---
+
+## Testing & Quality Assurance
+
+CATranS utilizes `pytest` for automated test suites, `ruff` for code style consistency, and `mypy` for strict type checking.
+
+```bash
+# Run tests
+uv run pytest
+
+# Check code format and style
+uv run ruff check .
+
+# Perform static type checks
+uv run mypy src/catrans
+```
+
+---
+
+## License
+
+This project is licensed under the MIT License - see the [LICENSE](LICENSE) file for details.

catrans-0.1.0/README.md

@@ -0,0 +1,241 @@
+# CATranS: Clinical Acute Transcriptomic Stratification
+
+**CATranS** is a robust bioinformatics and machine learning pipeline written in Python designed to stratify acute clinical cohorts using transcriptomic count data. 
+
+Through an iterative, semi-supervised learning approach, CATranS combines data normalization, batch correction, correlation-based feature selection, and class prediction to refine classifications of patient cohorts (e.g., separating acute from non-acute states, or predicting survival outcomes).
+
+---
+
+## Key Features
+
+- **Iterative Semi-Supervised Relabeling**: Safely expands training cohorts by predicting clinical labels on unclassified/unlabeled samples using high-confidence thresholds.
+- **Generalized Custom Classes**: Supports user-defined positive and negative class labeling (e.g., `Survivor`/`NonSurvivor`, `Mild`/`Severe`, `Responders`/`NonResponders`) with robust Categorical ordering to fix Wald contrast directions in DESeq2.
+- **Batch Correction**: Integrated with `inmoose` for ComBat-seq count batch effect correction.
+- **DESeq2-style Normalization**: Leverages the high-performance `pydeseq2` implementation for median-of-ratios normalization.
+- **Correlation-Based Feature Learning**: Custom graph-theory based gene selection utilizing `igraph`. Filters out highly redundant co-expressed genes and selects optimal representative biomarkers correlating with clinical outcomes.
+- **Flexible Estimator Integration**: Compatible with any Scikit-Learn classifier exposing a `predict_proba` method.
+
+---
+
+## Workflow Overview
+
+```mermaid
+graph TD
+    A[Raw Count Matrix & Metadata] --> B[Data Normalization & Optional ComBat-seq Batch Correction]
+    B --> C[Feature Selection: Spearman Correlation with Labels]
+    C --> D[Graph Construction: Pearson Gene-Gene Network]
+    D --> E[Redundant Feature Compression: Connected Components Vertex Selection]
+    E --> F[Iterative Semi-Supervised Model Training]
+    F --> G[Fit Classifier & Predict Unlabeled Samples]
+    G --> H{Meets Threshold?}
+    H -- Yes --> I[Relabel Samples & Run DESeq2 Differential Expression Analysis]
+    I --> J{Are there new DEGs?}
+    J -- Yes --> F
+    J -- No --> K[Final Output Classifications & Iteration History]
+    H -- No --> K
+```
+
+---
+
+## Installation
+
+We recommend using [uv](https://github.com/astral-sh/uv) or `pip` for rapid virtual environment setup and dependency management.
+
+```bash
+# Clone the repository
+git clone https://github.com/ruben-canas-ivi/CATranS.git
+cd CATranS
+
+# Install dependencies and the package using uv
+uv sync
+```
+
+### Dependencies
+- Python `>=3.11`
+- `numpy>=2.3.0`
+- `pandas>=2.3.3`
+- `scipy>=1.14.0`
+- `scikit-learn>=1.8.0`
+- `pydeseq2>=0.5.4`
+- `igraph>=1.0.0`
+- `inmoose>=0.9.1`
+
+---
+
+## API Reference & Usage Scenarios
+
+### Core Classes & Functions
+
+#### 1. `CatransClasses`
+Defines the names of the binary classes.
+- `positive` (default `"Dead"`): The target/case class.
+- `negative` (default `"Alive"`): The reference/control class.
+
+#### 2. `correlation_based_feature_selection`
+Identifies representative biomarker genes by correlating expression levels with labels, building a gene-gene correlation network, and selecting one representative gene per connected component.
+
+#### 3. `run_catrans`
+Runs the iterative semi-supervised relabeling pipeline.
+- `raw_counts` (pd.DataFrame): Genes (rows) by Samples (columns).
+- `metadata` (pd.DataFrame): Samples (rows) with metadata.
+- `acute_classification` (pd.DataFrame): Samples (rows) with initially known labels under the column `"class"` (unlabeled samples should be `NaN`).
+- `model`: Any Scikit-Learn compatible classifier (e.g. `RandomForestClassifier`, `SVC(probability=True)`).
+- `classes` (CatransClasses, Optional): Custom positive/negative class names.
+- `batch_col` (str, Optional): Metadata column to use for ComBat-seq batch correction.
+- `model_threshold` (float, default `0.8`): Confidence threshold to assign a label.
+- `signature_genes` (list[str], Optional): Genes to use for training the model. If `None`, `correlation_based_feature_selection` runs automatically.
+
+---
+
+### Usage Scenarios
+
+#### Scenario 1: Normal Workflow (Two-step Feature Selection + Stratification)
+This is the recommended workflow where you first find a non-redundant gene signature, then pass it to the CATranS pipeline.
+
+```python
+import pandas as pd
+from sklearn.ensemble import RandomForestClassifier
+import catrans
+
+# 1. Load your transcriptomics and clinical datasets
+counts = pd.read_csv("data/raw_counts.tsv", sep="\t", index_col=0)
+metadata = pd.read_csv("data/metadata.tsv", sep="\t", index_col=0)
+initial_classes = pd.read_csv("data/initial_classes.tsv", sep="\t", index_col=0)
+
+# 2. Define custom labels: "Survivor" (negative) and "NonSurvivor" (positive)
+my_classes = catrans.CatransClasses(positive="NonSurvivor", negative="Survivor")
+
+# 3. Perform correlation-based feature selection to identify candidate biomarkers
+signature = catrans.correlation_based_feature_selection(
+    data=counts,
+    acute_classification=initial_classes,
+    classes=my_classes,
+    pval=0.05,
+    cutoff_corr=0.5,
+    cutoff_corr_genes=0.8,
+    n_cpus=4
+)
+
+# 4. Run the semi-supervised relabeling pipeline using the selected signature
+results = catrans.run_catrans(
+    raw_counts=counts,
+    metadata=metadata,
+    acute_classification=initial_classes,
+    classes=my_classes,
+    model=RandomForestClassifier(random_state=42),
+    model_threshold=0.75,
+    signature_genes=signature,
+    n_cpus=4
+)
+
+print(results.final_classification)
+```
+
+#### Scenario 2: Predefined/Custom Gene Signature (Direct Stratification)
+If you already have a target gene signature (e.g., from literature or a previous experiment), you can bypass the feature selection step entirely and run the stratification pipeline directly.
+
+```python
+import pandas as pd
+from sklearn.svm import SVC
+import catrans
+
+counts = pd.read_csv("data/raw_counts.tsv", sep="\t", index_col=0)
+metadata = pd.read_csv("data/metadata.tsv", sep="\t", index_col=0)
+initial_classes = pd.read_csv("data/initial_classes.tsv", sep="\t", index_col=0)
+
+# Custom predefined signature from literature
+custom_signature = ["ENSG00000034693.15", "ENSG00000101188.5", "ENSG00000104870.13"]
+
+# Run the CATranS pipeline directly with the custom signature
+results = catrans.run_catrans(
+    raw_counts=counts,
+    metadata=metadata,
+    acute_classification=initial_classes,
+    model=SVC(probability=True, random_state=42),
+    model_threshold=0.8,
+    signature_genes=custom_signature,
+    n_cpus=4
+)
+
+print(f"Completed in {results.number_of_iterations} iterations.")
+```
+
+#### Scenario 3: Implicit Feature Selection (Defaulting Signature Genes)
+If you do not specify a list of `signature_genes`, `run_catrans` will automatically execute `correlation_based_feature_selection` under the hood using default parameters.
+
+```python
+import pandas as pd
+from sklearn.ensemble import RandomForestClassifier
+import catrans
+
+counts = pd.read_csv("data/raw_counts.tsv", sep="\t", index_col=0)
+metadata = pd.read_csv("data/metadata.tsv", sep="\t", index_col=0)
+initial_classes = pd.read_csv("data/initial_classes.tsv", sep="\t", index_col=0)
+
+# Run CATranS (it will automatically perform feature selection internally)
+results = catrans.run_catrans(
+    raw_counts=counts,
+    metadata=metadata,
+    acute_classification=initial_classes,
+    model=RandomForestClassifier(random_state=42),
+    model_threshold=0.75,
+    n_cpus=4
+)
+```
+
+#### Scenario 4: Default Thresholds & Core Settings
+You can omit optional arguments like `model_threshold` or `classes` if you wish to use the defaults (`model_threshold=0.8`, `classes` defaulting to `"Dead"`/`"Alive"`).
+
+```python
+import pandas as pd
+from sklearn.ensemble import RandomForestClassifier
+import catrans
+
+counts = pd.read_csv("data/raw_counts.tsv", sep="\t", index_col=0)
+metadata = pd.read_csv("data/metadata.tsv", sep="\t", index_col=0)
+initial_classes = pd.read_csv("data/initial_classes.tsv", sep="\t", index_col=0)
+
+# Minimal parameters
+results = catrans.run_catrans(
+    raw_counts=counts,
+    metadata=metadata,
+    acute_classification=initial_classes,
+    model=RandomForestClassifier(random_state=42)
+)
+```
+
+#### Scenario 5: Running with Batch Correction
+To normalize and batch-correct raw counts concurrently, specify the name of the batch column in your metadata dataframe:
+
+```python
+results = catrans.run_catrans(
+    raw_counts=counts,
+    metadata=metadata,
+    acute_classification=initial_classes,
+    model=RandomForestClassifier(random_state=42),
+    batch_col="batch"  # ComBat-seq batch effect correction column
+)
+```
+
+---
+
+## Testing & Quality Assurance
+
+CATranS utilizes `pytest` for automated test suites, `ruff` for code style consistency, and `mypy` for strict type checking.
+
+```bash
+# Run tests
+uv run pytest
+
+# Check code format and style
+uv run ruff check .
+
+# Perform static type checks
+uv run mypy src/catrans
+```
+
+---
+
+## License
+
+This project is licensed under the MIT License - see the [LICENSE](LICENSE) file for details.

catrans-0.1.0/pyproject.toml

@@ -0,0 +1,67 @@
+[project]
+name = "CATranS"
+version = "0.1.0"
+description = "Clinical Acute Transcriptomic Stratification"
+authors = [
+    { name = "Patricia Sebastián-León", email = "patricia.sebastian@ivirma.com" },
+    { name = "Antonio Parraga-Leo",     email = "antonio.parraga@ivirma.com" },
+    { name = "Rubén Cañas-Cañas",       email = "ruben.canas@ivirma.com" }
+]
+readme = "README.md"
+requires-python = ">=3.11"
+keywords = ["bioinformatics", "machine-learning", "deseq2", "stratification", "genomics"]
+classifiers = [
+    "Development Status :: 4 - Beta",
+    "Intended Audience :: Science/Research",
+    "Topic :: Scientific/Engineering :: Bio-Informatics",
+    "License :: OSI Approved :: MIT License",
+    "Programming Language :: Python :: 3",
+    "Programming Language :: Python :: 3.11",
+    "Programming Language :: Python :: 3.12",
+    "Operating System :: OS Independent",
+]
+
+dependencies = [
+    "numpy>=2.3.0",
+    "pandas>=2.3.3",
+    "scikit-learn>=1.8.0",
+    "scipy>=1.17.1",
+    "pydeseq2>=0.5.4",
+    "igraph>=1.0.0",
+    "inmoose>=0.9.1",
+]
+
+[project.urls]
+Homepage = "https://github.com/ruben-canas-ivi/CATranS"
+Issues = "https://github.com/ruben-canas-ivi/CATranS/issues"
+
+[build-system]
+requires = ["hatchling"]
+build-backend = "hatchling.build"
+
+[dependency-groups]
+dev = [
+    "pytest>=8.0.0",
+    "ruff>=0.9.0",
+    "mypy>=1.14.0",
+]
+
+[tool.ruff]
+line-length = 120
+target-version = "py311"
+
+[tool.mypy]
+python_version = "3.11"
+strict = true
+
+[[tool.mypy.overrides]]
+module = [
+    "pydeseq2.*",
+    "igraph.*",
+    "inmoose.*",
+    "statsmodels.*",
+    "scipy.*",
+    "pandas.*"
+]
+ignore_missing_imports = true
+

catrans-0.1.0/src/catrans/__init__.py

@@ -0,0 +1,14 @@
+from catrans.core import (
+    correlation_based_feature_selection,
+    run_catrans,
+    _diffExpAnalysis # recordar quitar esto antes de subir el paquete, que esto es solo para validaciones #TODO
+    )
+from catrans.data import CatransClasses, CatransResult
+
+__all__ = [
+    'correlation_based_feature_selection',
+    'run_catrans',
+    '_diffExpAnalysis', # recordar quitar esto antes de subir el paquete, que esto es solo para validaciones #TODO
+    'CatransClasses',
+    'CatransResult',
+    ]

catrans-0.1.0/src/catrans/_logging.py

@@ -0,0 +1,38 @@
+import logging
+import sys
+
+logger = logging.getLogger(__name__)
+logger.propagate = False
+
+class CustomFormatter(logging.Formatter):
+    info_format = "[%(asctime)s] | %(levelname)-8s | %(message)s"
+    debug_format = "[%(asctime)s] | %(levelname)-8s | STEP: %(funcName)s | %(message)s"
+
+    def format(self, record: logging.LogRecord) -> str:
+        orig_levelname = record.levelname
+        record.levelname = orig_levelname.center(8)
+       
+        if record.levelno == logging.DEBUG:
+            self._style._fmt = self.debug_format
+        else:
+            self._style._fmt = self.info_format
+        return super().format(record)
+
+def _get_logger() -> logging.Logger:
+    return logging.getLogger(name=__name__)
+
+def _setup_internal_logger(verbose: bool) -> None:
+    # Si no hay handlers, creamos uno con el CustomFormatter
+    if not logger.handlers:
+        handler = logging.StreamHandler(sys.stdout)
+        handler.setFormatter(CustomFormatter(datefmt='%y-%m-%d %H:%M:%S'))
+        logger.addHandler(handler)
+    
+    # Si el usuario activa verbose, bajamos a INFO
+    # Si tú desde fuera pusiste DEBUG, esto no lo sobrescribe (porque DEBUG < INFO)
+    if verbose:
+        # if logger.level > logging.INFO or logger.level == logging.NOTSET:
+        logger.setLevel(logging.INFO)
+    else:
+        if logger.level == logging.NOTSET or logger.level == logging.INFO:
+            logger.setLevel(logging.WARNING)