mathLab / PyDMD / 205
98%

DEFAULT BRANCH: master
Ran
Jobs 4
Files 19
Run time 6min
Badge
Embed ▾
README BADGES
x

If you need to use a raster PNG badge, change the '.svg' to '.png' in the link

Markdown

Textile

RDoc

HTML

Rst
205

push

travis-ci

mtezzele 
Low-rank DMD: Optimal solution in poylnomial time (#87)

* Implementation of the optimal closed-form DMD solution

- Added a predict function following the sklearn API.

* Added Tutorial 8 (Rough sketch)

- Added a tutorial comparing the performances of the various DMD implementations.

* Added `predict` to DMD.

- Added a `predict` function to the DMD estimator.

* Clean-ish implementation of the algorithm by Héas & Herzet

* Store the singular value of the low-dimensional DMD operator into self._svds.

* Change default DMD-operator factorization to EVD instead of SVD.

* Added Module DocString.

* Change pinv for a more efficient implementation

- Added an implementation of pseudo inverse computation dedicated to diagonal matrix. More efficient for this particular task than calling the general-purpose scipy.linalg.pinv.

- Implemented `_buiild_lowrank_op` and `_eig_from_lowrank_op` for consistency with the other DMD objects available in PyDMD.

- Sketched `_compute_amplitude` and `dynamics` for future implementation.

- Added basic tests for the current implementation of OptDMD. Note that it still misses a few functionalities that cannot hence be tested yet.

* Added optdmd to the __all__

* Remove super().__init__() for python 2/3 compatibility.

* Modified the test for atilde

- DMD and OptDMD construct slightly different low-dimensional DMD operators. The eigenvalues should be the same however.

1549 of 1579 relevant lines covered (98.1%)

3.92 hits per line

Jobs
ID Job ID Ran Files Coverage
1 205.1 (TOXENV=py27) 0
98.1
 Travis Job 205.1
2 205.2 (TOXENV=py36) 0
98.1
 Travis Job 205.2
3 205.3 (TOXENV=py27) 0
98.1
 Travis Job 205.3
4 205.4 (TOXENV=py36) 0
98.1
 Travis Job 205.4
Source Files on build 205
Detailed source file information is not available for this build.