Merge pull request #35 from iomega/refactor_environment

Refactor to work with matchms 0.6.0
iomega · Sep 16, 2020 · f2a7253 · f2a7253
2 parents 7f2cc8c + bbac633
commit f2a7253
Show file tree

Hide file tree

Showing 15 changed files with 141 additions and 210 deletions.
diff --git a/.github/workflows/conda_build.yml b/.github/workflows/conda_build.yml
@@ -1,6 +1,9 @@
 name: Anaconda Build
 
-on: [push, pull_request]
+on:
+  push:
+  pull_request:
+    types: [opened, reopened]
 
 jobs:
 
@@ -11,7 +14,10 @@ jobs:
       fail-fast: false
       matrix:
         os: ['ubuntu-latest', 'macos-latest', 'windows-latest']
-        python-version: ["3.7"]
+        python-version: ["3.8"]
+        include:
+          - os: 'ubuntu-latest'
+            python-version: "3.7"
     steps:
       - uses: actions/checkout@v2
         with:
@@ -73,7 +79,7 @@ jobs:
         if: matrix.os == 'ubuntu-latest'
         run: sed -i "s+$PWD/++g" coverage.xml
       - name: SonarCloud Scan
-        if: matrix.os == 'ubuntu-latest'
+        if: ${{ matrix.os == 'ubuntu-latest' && matrix.python-version == '3.8' }}
         uses: sonarsource/sonarcloud-github-action@master
         env:
           GITHUB_TOKEN: ${{ secrets.GITHUB_TOKEN }}
@@ -123,7 +129,7 @@ jobs:
           [ "$RUNNING_OS" = "Windows" ] && export BUILDDIR=$RUNNER_TEMP\\spec2vec\\_build\\
           conda config --set anaconda_upload no
           conda build --numpy 1.18.1 --no-include-recipe \
-            --channel bioconda --channel conda-forge \
+            --channel nlesc --channel bioconda --channel conda-forge \
             --croot ${BUILDDIR} \
             ./conda
       - name: Upload package artifact from build
@@ -139,7 +145,7 @@ jobs:
       fail-fast: false
       matrix:
         os: ['ubuntu-latest', 'macos-latest', 'windows-latest']
-        python-version: ['3.7']
+        python-version: ['3.7', '3.8']
     runs-on: ${{ matrix.os }}
     needs: build
     steps:
@@ -194,7 +200,13 @@ jobs:
           conda install \
             --channel bioconda \
             --channel conda-forge \
-            --channel $BUILDDIR \
+            --channel nlesc \
+            matchms
+          # Install spec2vec without nlesc channel to prevent package to be installed from nlesc channel instead of wanted $BUILDDIR channel 
+          conda install \
+            --channel bioconda \
+            --channel conda-forge \
+            --channel $BUILDDIR -v \
             spec2vec
       - name: List conda packages
         shell: bash -l {0}

diff --git a/.github/workflows/conda_publish.yml b/.github/workflows/conda_publish.yml
@@ -36,7 +36,9 @@ jobs:
           export BUILD_FOLDER=/tmp/spec2vec/_build
           mkdir -p $BUILD_FOLDER
           conda build --numpy 1.18.1 --no-include-recipe \
-            --channel bioconda --channel conda-forge \
+            --channel nlesc \
+            --channel bioconda \
+            --channel conda-forge \
             --croot $BUILD_FOLDER \
             ./conda
       - name: Push the package to anaconda cloud

diff --git a/CHANGELOG.md b/CHANGELOG.md
@@ -7,6 +7,18 @@ and this project adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0
 
 ## [Unreleased]
 
+### Added
+
+- Support for Python 3.8 [#35](https://github.com/iomega/spec2vec/pull/35)
+
+### Changed
+
+- Refactored Spec2Vec class to provide .pair() and .matrix() methods [#35](https://github.com/iomega/spec2vec/pull/35)
+
+### Removed
+
+- Spec2VecParallel (is now included as Spec2Vec.matrix()) [#35](https://github.com/iomega/spec2vec/pull/35)
+
 ## [0.2.0] - 2020-06-18
 
 ### Added

diff --git a/README.rst b/README.rst
@@ -99,14 +99,14 @@ Installation
 
 Prerequisites:  
 
-- Python 3.7  
+- Python 3.7 or 3.8  
 - Anaconda
 
 Install spec2vec from Anaconda Cloud with
 
 .. code-block:: console
 
-  conda env create --name spec2vec python=3.7
+  conda env create --name spec2vec python=3.8
   conda activate spec2vec
   conda install --channel nlesc --channel bioconda --channel conda-forge spec2vec
 

diff --git a/conda/environment-build.yml b/conda/environment-build.yml
@@ -5,4 +5,4 @@ dependencies:
   - anaconda-client
   - conda-build
   - conda-verify
-  - python >=3.7,<3.8
+  - python >=3.7,<3.9
diff --git a/conda/environment-dev.yml b/conda/environment-dev.yml
@@ -6,10 +6,10 @@ channels:
   - nlesc
 dependencies:
   - gensim >=3.8.0
-  - matchms>=0.5.0
+  - matchms>=0.6.0
   - numpy
   - pip
-  - python >=3.7,<3.8
+  - python >=3.7,<3.9
   - scipy
   - pip:
     - bump2version

diff --git a/conda/environment.yml b/conda/environment.yml
@@ -5,6 +5,7 @@ channels:
   - defaults
 dependencies:
   - gensim >=3.8.0
+  - matchms >=0.6.0
   - numpy
-  - python >=3.7,<3.8
+  - python >=3.7,<3.9
   - scipy
diff --git a/conda/meta.yaml b/conda/meta.yaml
@@ -10,6 +10,7 @@ source:
 
 extra:
   channels:
+    - nlesc
     - conda-forge
     - bioconda
 
@@ -26,18 +27,20 @@ requirements:
     - conda-verify
     - pytest-runner
     - python
+    - matchms >=0.6.0
     - numpy {{ numpy }}
     - setuptools
   host:
-    - python >=3.7,<3.8
+    - python >=3.7,<3.9
     - pip
     - pytest-runner
     - setuptools
   run:
     - gensim >=3.8.0
+    - matchms >=0.6.0
     - numpy
     - pip
-    - python >=3.7,<3.8
+    - python >=3.7,<3.9
     - scipy
 
 test:

diff --git a/integration-tests/test_user_workflow_spec2vec_parallel.py b/integration-tests/test_user_workflow_spec2vec_parallel.py
diff --git a/readthedocs/index.rst b/readthedocs/index.rst
@@ -19,15 +19,15 @@ Installation
 
 Prerequisites:  
 
-- Python 3.7  
+- Python 3.7 or 3.8  
 - Anaconda
 
 Install spec2vec from Anaconda Cloud with
 
 .. code-block:: console
 
   # install spec2vec in a new virtual environment to avoid dependency clashes
-  conda create --name spec2vec python=3.7
+  conda create --name spec2vec python=3.8
   conda activate spec2vec
   conda install --channel nlesc --channel bioconda --channel conda-forge spec2vec
 

diff --git a/setup.py b/setup.py
@@ -39,7 +39,8 @@
         "License :: OSI Approved :: Apache Software License",
         "Natural Language :: English",
         "Programming Language :: Python :: 3",
-        "Programming Language :: Python :: 3.7"
+        "Programming Language :: Python :: 3.7",
+        "Programming Language :: Python :: 3.8"
     ],
     test_suite="tests",
     install_requires=[

diff --git a/spec2vec/Spec2Vec.py b/spec2vec/Spec2Vec.py
@@ -1,11 +1,14 @@
+from typing import List
 from typing import Union
+import numpy
 import scipy
 from gensim.models.basemodel import BaseTopicModel
+from matchms.similarity.BaseSimilarity import BaseSimilarity
 from spec2vec.SpectrumDocument import SpectrumDocument
 from .calc_vector import calc_vector
 
 
-class Spec2Vec:
+class Spec2Vec(BaseSimilarity):
     """Calculate spec2vec similarity scores between a reference and a query.
 
     Using a trained model, spectrum documents will be converted into spectrum
@@ -64,7 +67,7 @@ def __init__(self, model: BaseTopicModel, intensity_weighting_power: Union[float
         self.allowed_missing_percentage = allowed_missing_percentage
         self.vector_size = model.wv.vector_size
 
-    def __call__(self, reference: SpectrumDocument, query: SpectrumDocument) -> float:
+    def pair(self, reference: SpectrumDocument, query: SpectrumDocument) -> float:
         """Calculate the spec2vec similaritiy between a reference and a query.
 
         Parameters
@@ -86,3 +89,38 @@ def __call__(self, reference: SpectrumDocument, query: SpectrumDocument) -> floa
         cdist = scipy.spatial.distance.cosine(reference_vector, query_vector)
 
         return 1 - cdist
+
+    def matrix(self, references: List[SpectrumDocument], queries: List[SpectrumDocument],
+               is_symmetric: bool = False) -> numpy.ndarray:
+        """Calculate the spec2vec similarities between all references and queries.
+
+        Parameters
+        ----------
+        references:
+            Reference spectrum documents.
+        queries:
+            Query spectrum documents.
+
+        Returns
+        -------
+        spec2vec_similarity
+            Array of spec2vec similarity scores.
+        """
+        n_rows = len(references)
+        reference_vectors = numpy.empty((n_rows, self.vector_size), dtype="float")
+        for index_reference, reference in enumerate(references):
+            reference_vectors[index_reference, 0:self.vector_size] = calc_vector(self.model,
+                                                                                 reference,
+                                                                                 self.intensity_weighting_power,
+                                                                                 self.allowed_missing_percentage)
+        n_cols = len(queries)
+        query_vectors = numpy.empty((n_cols, self.vector_size), dtype="float")
+        for index_query, query in enumerate(queries):
+            query_vectors[index_query, 0:self.vector_size] = calc_vector(self.model,
+                                                                         query,
+                                                                         self.intensity_weighting_power,
+                                                                         self.allowed_missing_percentage)
+
+        spec2vec_similarity = 1 - scipy.spatial.distance.cdist(reference_vectors, query_vectors, "cosine")
+
+        return spec2vec_similarity