Merge pull request #120 from MannLabs/make-dask-optional

mschwoer · web-flow · commit c4e3c9b78642 · 2026-01-28T17:21:35.000+01:00
Make dask optional
diff --git a/.binder/Dockerfile b/.binder/Dockerfile
@@ -27,7 +27,7 @@ COPY LICENSE LICENSE
 COPY README.md README.md
 COPY pyproject.toml pyproject.toml
 
-RUN pip install ".[stable]"
+RUN pip install ".[stable,dask-stable]"
 
 # change owner to be able to write to that folder
 COPY --chown=${USER}:${USER} example_nbs example_nbs
diff --git a/.github/workflows/e2e_tests_quick.yml b/.github/workflows/e2e_tests_quick.yml
@@ -28,5 +28,5 @@ jobs:
     with:
       python-version: ${{ matrix.python-version }}
       os: ${{ matrix.os }}
-      install-script: ./pip_install.sh tests,gui
+      install-script: ./pip_install.sh tests,gui,dask
       test-script: ./run_e2e_tests_quick.sh
diff --git a/.github/workflows/e2e_tests_quick_multiple_platforms.yml b/.github/workflows/e2e_tests_quick_multiple_platforms.yml
@@ -29,7 +29,7 @@ jobs:
     with:
       python-version: ${{ matrix.python-version }}
       os: ${{ matrix.os }}
-      install-script: ./pip_install.sh stable,tests,gui
+      install-script: ./pip_install.sh stable,tests,gui,dask-stable
       test-script: ./run_e2e_tests_quick.sh
 
   run-unit-tests-loose:
@@ -43,5 +43,5 @@ jobs:
     with:
       python-version: ${{ matrix.python-version }}
       os: ${{ matrix.os }}
-      install-script: ./pip_install.sh tests,gui
+      install-script: ./pip_install.sh tests,gui,dask
       test-script: ./run_e2e_tests_quick.sh
diff --git a/.github/workflows/install_and_unit_tests.yml b/.github/workflows/install_and_unit_tests.yml
@@ -30,5 +30,5 @@ jobs:
     with:
       python-version: ${{ matrix.python-version }}
       os: ${{ matrix.os }}
-      install-script: ./pip_install.sh tests,gui
+      install-script: ./pip_install.sh tests,gui,dask
       test-script: ./run_unit_tests.sh
diff --git a/.github/workflows/install_and_unit_tests_multiple_platforms.yml b/.github/workflows/install_and_unit_tests_multiple_platforms.yml
@@ -26,7 +26,7 @@ jobs:
     with:
       python-version: ${{ matrix.python-version }}
       os: ${{ matrix.os }}
-      install-script: ./pip_install.sh stable,tests,gui
+      install-script: ./pip_install.sh stable,tests,gui,dask-stable
       test-script: ./run_unit_tests.sh
 
   run-unit-tests-loose:
@@ -40,5 +40,5 @@ jobs:
     with:
       python-version: ${{ matrix.python-version }}
       os: ${{ matrix.os }}
-      install-script: ./pip_install.sh tests,gui
+      install-script: ./pip_install.sh tests,gui,dask
       test-script: ./run_unit_tests.sh
diff --git a/.github/workflows/run_example_nbs.yml b/.github/workflows/run_example_nbs.yml
@@ -28,5 +28,5 @@ jobs:
     with:
       python-version: ${{ matrix.python-version }}
       os: ${{ matrix.os }}
-      install-script: ./pip_install.sh tests,gui
+      install-script: ./pip_install.sh tests,gui,dask
       test-script: ./run_example_notebooks.sh
diff --git a/Dockerfile b/Dockerfile
@@ -20,7 +20,7 @@ COPY LICENSE LICENSE
 COPY README.md README.md
 COPY pyproject.toml pyproject.toml
 
-RUN pip install ".[stable,gui-stable]"
+RUN pip install ".[stable,gui-stable,dask-stable]"
 
 ENV PORT=41215
 EXPOSE 41215
diff --git a/README.md b/README.md
@@ -102,6 +102,18 @@ For those who are really adventurous, it is also possible to directly install an
 pip install "git+https://github.com/MannLabs/alphaquant.git@development#egg=alphaquant[stable,development-stable]"
 ```
 
+#### Optional extras
+
+For processing very large PTM files that don't fit in memory, install the optional dask extra:
+
+```bash
+pip install "alphaquant[dask]"
+```
+or the version with a pinned dependency (recommended)
+```bash
+pip install "alphaquant[dask-stable]"
+```
+
 ### Developer installation
 
 AlphaQuant can also be installed in editable (i.e. developer) mode with a few `bash` commands. This allows to fully customize the software and even modify the source code to your specific needs. When an editable Python package is installed, its source code is stored in a transparent location of your choice. While optional, it is advised to first (create and) navigate to e.g. a general software folder:
diff --git a/alphaquant/ptm/ptmsite_mapping.py b/alphaquant/ptm/ptmsite_mapping.py
@@ -18,8 +18,16 @@
 # Cell
 
 import pandas as pd
-import dask.dataframe as dd
+import warnings
 
+try:
+    import dask.dataframe as dd
+    HAS_DASK = True
+except ModuleNotFoundError:
+    HAS_DASK = False
+    warnings.warn(
+        "Dependency 'dask' not installed. If you want to use its functionality, install it with: `pip install \"alphaquant[dask]\"` . ImportError will be raised if the data requires out-of-memory processing."
+    )
 
 
 def assign_dataset_chunkwise(input_file, results_dir, samplemap_df , modification_type = "[Phospho (STY)]", id_thresh = 0.6, excl_thresh =0.2 ,swissprot_file = None,
@@ -46,6 +54,12 @@ def assign_dataset_chunkwise(input_file, results_dir, samplemap_df , modificatio
         input_type (str): Type of input data ("Spectronaut" or "DIANN")
         organism (str): Organism name
     """
+    if not HAS_DASK:
+        raise ImportError(
+            "Dask is required for out-of-memory PTM site mapping. "
+            "Install it with: pip install \"alphaquant[dask]\""
+        )
+
     clean_up_previous_processings(results_dir)
 
     if input_type == 'Spectronaut':
diff --git a/alphaquant/quant_reader/longformat_reader.py b/alphaquant/quant_reader/longformat_reader.py
@@ -10,7 +10,7 @@
     HAS_DASK = True
 except ModuleNotFoundError:
     warnings.warn(
-        "Dependency 'dask' not installed. If you want to use its functionality, install alphaquant with the 'dask' extra. Falling back to non-dask based processing."
+        "Dependency 'dask' not installed. If you want to use its functionality, install it with: `pip install \"alphaquant[dask]\"` . Falling back to non-dask based processing."
     )
     HAS_DASK = False
 
diff --git a/pyproject.toml b/pyproject.toml
@@ -53,6 +53,8 @@ optional-dependencies = { stable = { file = ["requirements/requirements.txt",
 ] }, gui-stable = { file = [ "requirements/requirements_gui.txt",
 ] }, tests = { file = [ "requirements/requirements_tests.txt",
 ] }, development = { file = ["requirements/requirements_development.txt", "requirements/requirements_tests.txt",
+] }, dask = { file = [ "requirements/requirements_dask_loose.txt",
+] }, dask-stable = { file = [ "requirements/requirements_dask.txt",
 ] }}
 
 version = {attr = "alphaquant.__version__"}
diff --git a/release/linux/build_installer_linux.sh b/release/linux/build_installer_linux.sh
@@ -7,9 +7,8 @@ rm -rf dist_pyinstaller build_pyinstaller
 
 # Find the wheel file in dist directory
 WHL_NAME=$(cd dist && ls ./*.whl && cd ..)
-pip install "dist/${WHL_NAME}[stable,gui-stable]"
+pip install "dist/${WHL_NAME}[stable,gui-stable,dask-stable]"
 
 # Creating the stand-alone pyinstaller folder
-pip install pyinstaller
 pyinstaller release/pyinstaller/alphaquant.spec --distpath dist_pyinstaller --workpath build_pyinstaller -y
 
diff --git a/release/macos/build_installer_macos.sh b/release/macos/build_installer_macos.sh
@@ -7,9 +7,8 @@ rm -rf dist_pyinstaller build_pyinstaller
 
 # Get the wheel file name from dist directory
 WHL_NAME=$(cd dist && ls ./*.whl && cd ..)
-pip install "dist/${WHL_NAME}[stable,gui-stable]"
+pip install "dist/${WHL_NAME}[stable,gui-stable,dask-stable]"
 
 # Creating the stand-alone pyinstaller folder
-pip install pyinstaller
 pyinstaller release/pyinstaller/alphaquant.spec --distpath dist_pyinstaller --workpath build_pyinstaller -y
 
diff --git a/release/windows/build_installer_windows.ps1 b/release/windows/build_installer_windows.ps1
@@ -7,8 +7,7 @@ Remove-Item -Recurse -Force -ErrorAction SilentlyContinue ./dist_pyinstaller
 
 # substitute X.Y.Z-devN with X.Y.Z.devN
 $WHL_NAME = (Get-ChildItem -Path "dist" -Filter "*.whl").Name
-pip install "dist/$WHL_NAME[stable,gui-stable]"
+pip install "dist/$WHL_NAME[stable,gui-stable,dask-stable]"
 
 # Creating the stand-alone pyinstaller folder
-pip install pyinstaller
 pyinstaller release/pyinstaller/alphaquant.spec --distpath dist_pyinstaller --workpath build_pyinstaller -y
diff --git a/requirements/requirements.txt b/requirements/requirements.txt
@@ -9,9 +9,7 @@ anytree==2.12.1
 graphviz==0.20.3
 click==8.1.8
 lmfit==1.3.2
-dask==2025.1.0
 multiprocess==0.70.17
-pyfasta==0.5.2
 openpyxl==3.1.5
 scikit-learn==1.6.1
 alphamap==0.2.0
diff --git a/requirements/requirements_dask.txt b/requirements/requirements_dask.txt
@@ -0,0 +1 @@
+dask==2025.1.0
diff --git a/requirements/requirements_dask_loose.txt b/requirements/requirements_dask_loose.txt
@@ -0,0 +1 @@
+dask
diff --git a/requirements/requirements_loose.txt b/requirements/requirements_loose.txt
@@ -9,9 +9,7 @@ anytree
 graphviz
 click
 lmfit
-dask
 multiprocess
-pyfasta
 openpyxl
 scikit-learn
 alphabase

Original file line number	Diff line number	Diff line change
`@@ -10,7 +10,7 @@`
`10`	`10`	`HAS_DASK = True`
`11`	`11`	`except ModuleNotFoundError:`
`12`	`12`	`warnings.warn(`
`13`		`- "Dependency 'dask' not installed. If you want to use its functionality, install alphaquant with the 'dask' extra. Falling back to non-dask based processing."`
	`13`	+ "Dependency 'dask' not installed. If you want to use its functionality, install it with: `pip install \"alphaquant[dask]\"` . Falling back to non-dask based processing."
`14`	`14`	`)`
`15`	`15`	`HAS_DASK = False`
`16`	`16`