forked from peterwittek/somoclu
-
Notifications
You must be signed in to change notification settings - Fork 0
/
CHANGES
108 lines (88 loc) · 5.3 KB
/
CHANGES
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
Version 1.7.4: 2017-06-06
- Fixed: The random seed was set to 0 for testing purposes. This is now changed to a wall-time based initialization.
Version 1.7.3: 2017-04-21
- New: Verbosity parameter in the command-line, Python, MATLAB, and Julia interfaces.
- Changed: Calculation of U-matrix parallelized.
- Changed: Moved feeding data to train method in the Python interface.
- Fixed: Sparse matrix reader made more robust.
- Fixed: Compatibility with kohonen 3 resolved.
- Fixed: Compatibility with Matplotlib 2 resolved.
Version 1.7.2: 2016-11-24
- New: The coefficient of the Gaussian neighborhood function exp(-||x-y||^2/(2*(coeff*radius)^2)) is now exposed in all interfaces as a parameter.
- New: `get_bmu` function in the Python interface to get the best matching units given an activation map.
- Changed: Updated PCA initialization in the Python interface to work with `sk-learn` 0.18 onwards.
- Changed: Radii can be float values.
- Fixed: Only positive values were written back to codebook during update.
- Fixed: Sparse data is read correctly when there are class labels.
Version 1.7.1: 2016-10-02
- Fixed: macOS build works again.
Version 1.7.0: 2016-09-30
- New: Julia interface is available (https://github.com/peterwittek/Somoclu.jl).
- New: Method `get_surface_state` of the `Somoclu` object in Python calculates the activation map for all data instances.
- New: Method `view_activation_map` of the `Somoclu` object in Python allows plotting the activation map for the training data instances or for a new data instance.
- New: Method `view_similarity_matrix` of the `Somoclu` object in Python visualizes the similarity matrix of data points according to their distance to the nodes in the map.
- Fixed: CRAN-friendliness improved.
Version 1.6.2: 2016-08-09
- Changed: In-place codebook updates when compiled without MPI. This improves update speed and substantially cuts memory use.
- Changed: Compatible with Visual Studio 15.
- Fixed: The BMUs returned after training were from before the last epoch. Now another round of BMU search is done.
- Fixed: Training can continue on the same data in the Python wrapper.
- Fixed: GPU memory allocation problem on Windows.
Version 1.6.1: 2016-02-22
- New: Option for PCA initialization is added to the Python interface.
- New: Clustering of the codebook with arbitrary clustering algorithm in scikit-learn is now possible in the Python interface.
Version 1.6: 2016-01-10
- New: R wrapper integrates with kohonen package.
- New: MATLAB wrapper integrates with soomtoolbox.
- New: Better handling of CUDA compilation in the Python interface.
- Changed: Throws an exception if GPU kernel is requested, but it was compiled without it. The earlier behaviour quietly defaulted to the CPU kernel.
Version 1.5.1: 2015-12-02
- New: Neighborhood function can be chosen between Gaussian and bubble.
- Fixed: R wrapper passes arrays with correct orientation.
- Fixed: `io.cpp` is no longer required in the wrappers. An exception is thrown when needed.
Version 1.5: 2015-09-30
- New: Python interface has visual capabilities.
- New: Option for hexagonal grid.
- New: Option for requesting compact support in updating the map.
- New: Python, R, and MATLAB interfaces now allow passing an initial codebook.
- Changed: Reduced memory use in calculating U-matrices.
- Changed: Build system rebuilt and simplified.
Version 1.4.1: 2015-01-28
- Better support for ICC.
- Faster code when compiling with GCC.
- Building instructions and documentation improved.
- Bug fixes: portability for R, using native R random number generator.
Version 1.4: 2014-09-05
- Better Windows support.
- Completed CUDA support for Python and R interfaces.
- Faster compilation by removing unnecessary flags for nvcc and support for CUDA 6.5.
- Bug fixes: R version no longer needs separate code.
Version 1.3.1: 2014-04-10
- Better I/O separation for the Python, R, and MATLAB interfaces.
- Initial Windows support through GCC on Windows.
- Bug fixes: major MPI initialization bug fixed.
Version 1.3: 2014-03-31
- Learning rate parameter included.
- Linear and exponential cooling strategies added for radius and learning rate.
- CLI interface made more user-friendly.
- Default radius depends on both X and Y of the map.
- Bug fixes: CUDA build without MPI, best matching unit passing without MPI,
coordinate order in best matching unit file.
- Python, R, and MATLAB interfaces added.
Version 1.2: 2013-12-17
- Massive improvements in OpenMP parallelization.
- MPI libraries are no longer mandatory.
- Best matching units are saved.
- Option for specifying an initial codebook for the map.
- ESOM .lrn input format added.
- Parsing of white-space characters corrected.
- Long-named command line switches for specifying SOM dimensions.
- Fine-grained control of which interim files to save across epochs
- Option in Makefile for building shared library.
Version 1.1.2: 2013-11-28
- Toroid maps were added.
- Initial radius is exposed as a parameter via the command line interface.
- Formats of codebook and U-matrix export are compatible with Databionic ESOM Tools for advanced visualisation.
- Bug fixes: codebook update with a compact support was removed, NaN entry no longer appears in U-matrices.
Version 1.0: 2013-05-14
- Initial release.