Merge pull request #243 from pavlin-policar/updates

Various fixes and improvements before release

Author: pavlin-policar

azure-pipelines-release.yml

@@ -38,6 +38,11 @@ jobs:
           python.version: '3.10'
           ciwb.build: cp310-macosx_universal2
           ciwb.archs: universal2
+        osx - python311:
+          image.name: 'macos-12'
+          python.version: '3.11'
+          ciwb.build: cp311-macosx_universal2
+          ciwb.archs: universal2
 
         windows - python37:
           image.name: 'windows-2019'
@@ -59,6 +64,11 @@ jobs:
           python.version: '3.10'
           ciwb.build: cp310-win_amd64
           ciwb.archs: AMD64
+        windows - python311:
+          image.name: 'windows-2019'
+          python.version: '3.11'
+          ciwb.build: cp311-win_amd64
+          ciwb.archs: AMD64
 
     steps:
       - task: UsePythonVersion@0
@@ -93,9 +103,10 @@ jobs:
 
       - script: pip install pynndescent
         displayName: 'Install optional dependencies - pynndescent'
-        condition: ne(variables['python.version'], '3.10')
 
       - script: pip install hnswlib
+        env:
+          HNSWLIB_NO_NATIVE: 1  # -march=native is not available on clang, so just disable it
         displayName: 'Install optional dependencies - hnswlib'
 
       - bash: pip install pytest
@@ -139,6 +150,9 @@ jobs:
         python310:
           python: '/opt/python/cp310-cp310/bin'
           python.version: '3.10'
+        python311:
+          python: '/opt/python/cp311-cp311/bin'
+          python.version: '3.11'
 
     container:
       image: quay.io/pypa/manylinux2014_x86_64:latest
@@ -173,7 +187,6 @@ jobs:
 
       - script: $(python)/pip install pynndescent
         displayName: 'Install optional dependencies - pynndescent'
-        condition: ne(variables['python.version'], '3.10')
 
       - script: $(python)/pip install hnswlib
         displayName: 'Install optional dependencies - hnswlib'
@@ -213,9 +226,9 @@ jobs:
 
     steps:
       - task: UsePythonVersion@0
-        displayName: 'Use Python 3.8'
+        displayName: 'Use Python 3.10'
         inputs:
-          versionSpec: '3.8'
+          versionSpec: '3.10'
 
       - script: python -m pip install build
         displayName: Installing build tools
@@ -237,7 +250,6 @@ jobs:
 
       - script: pip install pynndescent
         displayName: 'Install optional dependencies - pynndescent'
-        condition: ne(variables['python.version'], '3.10')
 
       - script: pip install hnswlib
         displayName: 'Install optional dependencies - hnswlib'

azure-pipelines.yml

@@ -25,6 +25,9 @@ jobs:
       linux-python310:
         image.name: 'ubuntu-latest'
         python.version: '3.10'
+      linux-python311:
+        image.name: 'ubuntu-latest'
+        python.version: '3.11'
       osx-python37:
         image.name: 'macos-12'
         python.version: '3.7'
@@ -37,6 +40,9 @@ jobs:
       osx-python310:
         image.name: 'macos-12'
         python.version: '3.10'
+      osx-python311:
+        image.name: 'macos-12'
+        python.version: '3.11'
       windows-python37:
         image.name: 'windows-2019'
         python.version: '3.7'
@@ -49,6 +55,9 @@ jobs:
       windows-python310:
         image.name: 'windows-2019'
         python.version: '3.10'
+      windows-python311:
+        image.name: 'windows-2019'
+        python.version: '3.11'
 
   steps:
   - task: UsePythonVersion@0
@@ -75,9 +84,10 @@ jobs:
 
   - script: pip install pynndescent
     displayName: 'Install optional dependencies - pynndescent'
-    condition: ne(variables['python.version'], '3.10')
 
   - script: pip install hnswlib
+    env:
+      HNSWLIB_NO_NATIVE: 1  # -march=native is not available on clang, so just disable it
     displayName: 'Install optional dependencies - hnswlib'
 
   # Since Python automatically adds `cwd` to `sys.path`, it's important we remove the local folder

docs/source/api/affinity.rst

@@ -2,5 +2,5 @@ Affinity
 ========
 
 .. automodule:: openTSNE.affinity
-    :members: Affinities, PerplexityBasedNN, MultiscaleMixture, Multiscale, FixedSigmaNN, Uniform
+    :members: Affinities, PerplexityBasedNN, MultiscaleMixture, Multiscale, FixedSigmaNN, Uniform, PrecomputedAffinities
     :undoc-members:

docs/source/api/initialization.rst

@@ -2,4 +2,4 @@ Initialization
 ==============
 
 .. automodule:: openTSNE.initialization
-    :members: pca, random
+    :members: pca, spectral, random, rescale, jitter

docs/source/index.rst

@@ -30,9 +30,14 @@ openTSNE is a modular Python implementation of t-Distributed Stochasitc Neighbor
 References
 ----------
 
-.. [1] Van Der Maaten, Laurens, and Hinton, Geoffrey. `“Visualizing data using t-SNE” <http://www.jmlr.org/papers/volume9/vandermaaten08a/vandermaaten08a.pdf>`__, Journal of Machine Learning Research (2008).
-.. [2] Poličar, Pavlin G., Martin Stražar, and Blaž Zupan. `“Embedding to Reference t-SNE Space Addresses Batch Effects in Single-Cell Classification” <https://www.biorxiv.org/content/10.1101/671404v1.abstract>`__, BioRxiv (2019).
-.. [3] Van Der Maaten, Laurens. `“Accelerating t-SNE using tree-based algorithms” <http://www.jmlr.org/papers/volume15/vandermaaten14a/vandermaaten14a.pdf>`__, Journal of Machine Learning Research (2014).
+.. [1] Van der Maaten, Laurens, and Hinton, Geoffrey. `“Visualizing data using t-SNE” <http://www.jmlr.org/papers/volume9/vandermaaten08a/vandermaaten08a.pdf>`__, Journal of Machine Learning Research (2008).
+
+.. [2] Poličar, Pavlin G., Martin Stražar, and Blaž Zupan. `“Embedding to Reference t-SNE Space Addresses Batch Effects in Single-Cell Classification” <https://link.springer.com/article/10.1007/s10994-021-06043-1>`__, Machine Learning (2021).
+
+.. [3] Van der Maaten, Laurens. `“Accelerating t-SNE using tree-based algorithms” <http://www.jmlr.org/papers/volume15/vandermaaten14a/vandermaaten14a.pdf>`__, Journal of Machine Learning Research (2014).
+
 .. [4] Linderman, George C., et al. `"Fast interpolation-based t-SNE for improved visualization of single-cell RNA-seq data" <https://www.nature.com/articles/s41592-018-0308-4>`__, Nature Methods (2019).
+
 .. [5] Kobak, Dmitry, and Berens, Philipp. `“The art of using t-SNE for single-cell transcriptomics” <https://www.nature.com/articles/s41467-019-13056-x>`__, Nature Communications (2019).
+
 .. [6] Macosko, Evan Z., et al. \ `“Highly parallel genome-wide expression profiling of individual cells using nanoliter droplets” <https://www.sciencedirect.com/science/article/pii/S0092867415005498>`__, Cell (2015).

docs/source/parameters.rst

@@ -74,8 +74,8 @@ ints_in_interval: float
 
 References
 ----------
-.. [1] Kobak, Dmitry, and Philipp Berens. "The art of using t-SNE for single-cell transcriptomics." bioRxiv (2018): 453449.
+.. [1] Kobak, Dmitry, and Berens, Philipp. `“The art of using t-SNE for single-cell transcriptomics” <https://www.nature.com/articles/s41467-019-13056-x>`__, Nature Communications (2019).
 
-.. [2] Linderman, George C., and Stefan Steinerberger. "Clustering with t-SNE, provably." arXiv preprint arXiv:1706.02582 (2017).
+.. [2] Linderman, George C., and Stefan Steinerberger. `“Clustering with t-SNE, provably.” <https://epubs.siam.org/doi/abs/10.1137/18M1216134>`__, SIAM Journal on Mathematics of Data Science (2019).
 
-.. [3] Jacobs, Robert A. "Increased rates of convergence through learning rate adaptation." Neural networks 1.4 (1988): 295-307.
+.. [3] Jacobs, Robert A. `"Increased rates of convergence through learning rate adaptation." <https://www.sciencedirect.com/science/article/abs/pii/0893608088900032>`__, Neural Networks (1988).

docs/source/tsne_algorithm.rst

@@ -204,8 +204,8 @@ UMAP, a recent and popular embedding technique for visualizing high dimensional
 References
 ----------
 
-.. [1] Maaten, Laurens van der, and Geoffrey Hinton. "Visualizing data using t-SNE." Journal of machine learning research 9.Nov (2008): 2579-2605.
+.. [1] Van der Maaten, Laurens, and Hinton, Geoffrey. `“Visualizing data using t-SNE” <http://www.jmlr.org/papers/volume9/vandermaaten08a/vandermaaten08a.pdf>`__, Journal of Machine Learning Research (2008).
 
-.. [2] Van Der Maaten, Laurens. "Accelerating t-SNE using tree-based algorithms." The Journal of Machine Learning Research 15.1 (2014): 3221-3245.
+.. [2] Van der Maaten, Laurens. `“Accelerating t-SNE using tree-based algorithms” <http://www.jmlr.org/papers/volume15/vandermaaten14a/vandermaaten14a.pdf>`__, Journal of Machine Learning Research (2014).
 
-.. [3] Linderman, George C., et al. "Efficient Algorithms for t-distributed Stochastic Neighborhood Embedding." arXiv preprint arXiv:1712.09005 (2017).
+.. [3] Linderman, George C., et al. `"Fast interpolation-based t-SNE for improved visualization of single-cell RNA-seq data" <https://www.nature.com/articles/s41592-018-0308-4>`__, Nature Methods (2019).