skmad-suite issueshttps://gitlab.lis-lab.fr/groups/skmad-suite/-/issues2020-06-04T11:31:21+02:00https://gitlab.lis-lab.fr/skmad-suite/tff2020/-/issues/2pytest: make a new docker file updating depedencies2020-06-04T11:31:21+02:00Valentin Emiyapytest: make a new docker file updating depedencieshttps://gitlab.lis-lab.fr/skmad-suite/yafe/-/issues/3Measures must be float values, allow more options?2020-05-25T10:59:37+02:00Valentin EmiyaMeasures must be float values, allow more options?Only `float` values are allowed in the measures/performance. In particular, there is no way to output:
* complex values
* strings
* collections (fixed or variable length)
If one is contrained to have single values, one option could b...Only `float` values are allowed in the measures/performance. In particular, there is no way to output:
* complex values
* strings
* collections (fixed or variable length)
If one is contrained to have single values, one option could be to allow any type and to convert collections to single values, e.g. using `numpy.tobytes()` for encoding and `numpy.frombuffer` for decoding.https://gitlab.lis-lab.fr/skmad-suite/pyteuf/-/issues/1AttributeError: 'TfData' object has no attribute 'shrink_mask'2019-03-13T17:52:23+01:00Valentin EmiyaAttributeError: 'TfData' object has no attribute 'shrink_mask'An undesirable error is generated in the example below, when calling `plt.imshow(Y)` where `Y` is a `TfData` (with real values). To be avoided?
```python
import numpy as np
import matplotlib.pyplot as plt
from ltfatpy import gspi
from py...An undesirable error is generated in the example below, when calling `plt.imshow(Y)` where `Y` is a `TfData` (with real values). To be avoided?
```python
import numpy as np
import matplotlib.pyplot as plt
from ltfatpy import gspi
from pyteuf import Stft
x, fs = gspi()
win_len = 2048
stft_params = {
'win_name': 'hann',
'win_len': win_len,
'hop': win_len // 2,
'n_bins': win_len,
'convention': 'bp',
'zero_pad_full_sig': False,
'param_constraint': 'fix',
}
stft = Stft(**stft_params)
X = stft.apply(x)
Y = np.abs(X)
plt.imshow(Y)
```https://gitlab.lis-lab.fr/skmad-suite/madarrays/-/issues/4Switch to scipy resample function if resampy not present2019-03-13T14:23:44+01:00Ronan HamonSwitch to scipy resample function if resampy not presentresampy is annoying to install: it depends on `numba`, that depends on `llvmlite`, that requires last version of `llvm` installed.
I suggest letting the choice to the method of resampling depending on the configuration of the environmen...resampy is annoying to install: it depends on `numba`, that depends on `llvmlite`, that requires last version of `llvm` installed.
I suggest letting the choice to the method of resampling depending on the configuration of the environment.
If `resampy` is present, use `resampy`, otherwise use `signal.signal.resample`.Valentin EmiyaValentin Emiyahttps://gitlab.lis-lab.fr/skmad-suite/yafe/-/issues/2Allow functions `measure` to have a subset of all the current parameters.2018-10-25T16:03:33+02:00Valentin EmiyaAllow functions `measure` to have a subset of all the current parameters.Add an `inspect` mechanism in method `run_task_by_id`, update examples and documentation.Add an `inspect` mechanism in method `run_task_by_id`, update examples and documentation.https://gitlab.lis-lab.fr/skmad-suite/yafe/-/issues/1Create a method that runs a task given some # params without storing the resu...2018-10-25T16:01:21+02:00Valentin EmiyaCreate a method that runs a task given some # params without storing the results in data_pathUseful for demos (with GUI).Useful for demos (with GUI).