|
| 1 | +SUSI: SUpervised Self-organIzing maps in Python |
| 2 | +=============================================== |
| 3 | + |
| 4 | +Python package for unsupervised and supervised self-organizing maps (SOM) |
| 5 | + |
| 6 | +Description |
| 7 | +----------- |
| 8 | + |
| 9 | +We present the SUSI package for Python. |
| 10 | +It includes a fully functional SOM for unsupervised and supervised tasks. |
| 11 | +The class structure is set up as follows: |
| 12 | + |
| 13 | +- SOMClustering: Unsupervised SOM for clustering |
| 14 | + |
| 15 | + - SOMEstimator: Base class for supervised SOMs |
| 16 | + |
| 17 | + - SOMRegressor: Regression SOM |
| 18 | + - SOMClassifier: Classification SOM |
| 19 | + |
| 20 | + |
| 21 | +:Paper: |
| 22 | + to be published |
| 23 | + |
| 24 | +:License: |
| 25 | + `3-Clause BSD license <LICENSE>`_ |
| 26 | + |
| 27 | +:Authors: |
| 28 | + `Felix M. Riese <mailto:github@felixriese.de>`_, |
| 29 | + `Sina Keller <mailto:sina.keller@kit.edu>`_ |
| 30 | + |
| 31 | +:Citation: see `Citation`_ and in the `bibtex <bibliography.bib>`_ file |
| 32 | + |
| 33 | +Installation |
| 34 | +------------ |
| 35 | + |
| 36 | +With PyPi: |
| 37 | + |
| 38 | +.. code:: bash |
| 39 | +
|
| 40 | + pip3 install susi |
| 41 | +
|
| 42 | +
|
| 43 | +Manually: |
| 44 | + |
| 45 | +.. code:: bash |
| 46 | +
|
| 47 | + git clone TODO |
| 48 | + cd susi/ |
| 49 | + python setup.py install |
| 50 | +
|
| 51 | +**Dependencies** |
| 52 | + |
| 53 | +Python 3 with: |
| 54 | + |
| 55 | +* joblib |
| 56 | +* numpy |
| 57 | +* scikit-learn |
| 58 | +* scipy |
| 59 | + |
| 60 | +Usage |
| 61 | +----- |
| 62 | + |
| 63 | +Regression in python3: |
| 64 | + |
| 65 | +.. code:: python3 |
| 66 | +
|
| 67 | + import susi |
| 68 | +
|
| 69 | + som = susi.SOMRegressor() |
| 70 | + som.fit(X_train, y_train) |
| 71 | + print(som.score(X_test, y_test)) |
| 72 | +
|
| 73 | +
|
| 74 | +Classification in `python3`: |
| 75 | + |
| 76 | +.. code:: python3 |
| 77 | +
|
| 78 | + import susi |
| 79 | +
|
| 80 | + som = susi.SOMClassifier() |
| 81 | + som.fit(X_train, y_train) |
| 82 | + print(som.score(X_test, y_test)) |
| 83 | +
|
| 84 | +Code examples as Jupyter Notebooks: |
| 85 | + |
| 86 | +* `examples/SOMRegressor_Hyperspectral <examples/SOMRegressor_Hyperspectral.ipynb>`_ |
| 87 | +* `examples/SOMClassifier <examples/SOMClassifier.ipynb>`_ |
| 88 | +* `examples/SOMClassifier_Salinas <examples/SOMClassifier_Salinas.ipynb>`_ |
| 89 | + |
| 90 | +Citation |
| 91 | +-------- |
| 92 | + |
| 93 | +The bibtex file including both references is available `here <bibliography.bib>`_. |
| 94 | + |
| 95 | +**Paper:** |
| 96 | + |
| 97 | +F. M. Riese and S. Keller, "SUSI: Supervised self-organizing maps for regression and classification in Python", 2019, Submitted to an ISPRS conference. |
| 98 | + |
| 99 | +.. code:: bibtex |
| 100 | +
|
| 101 | + @article{riese2019susi, |
| 102 | + author = {Riese, Felix~M. and Keller, Sina}, |
| 103 | + title = {SUSI: Supervised self-organizing maps for regression and classification in Python}, |
| 104 | + year = {2019}, |
| 105 | + notes = {Submitted to an ISPRS conference}, |
| 106 | + TODO arxiv |
| 107 | + } |
| 108 | +
|
| 109 | +
|
| 110 | +**Code:** |
| 111 | + |
| 112 | +F. M. Riese, "TODO", [DOI TODO](DOI TODO), 2019. |
| 113 | + |
| 114 | +TODO Badge |
| 115 | + |
| 116 | +.. code:: bibtex |
| 117 | +
|
| 118 | + @misc{riese2019cnn, |
| 119 | + author = {Riese, Felix~M.}, |
| 120 | + title = {{TODO}}, |
| 121 | + year = {2019}, |
| 122 | + publisher = {Zenodo}, |
| 123 | + DOI = {TODO}, |
| 124 | + howpublished = {\href{TODO}{TODO}} |
| 125 | + } |
0 commit comments