-
This is a first step towards the broader goal of running the workflow with all new species. Currently there are 5 new species datasets on Movebank and 2 more that are not on Movebank.
Received app…
-
I am running a study in EEG lab. I wish to plot ERSP and ITC plots but i am not able to do so in times vs frequency plots. On the other hand i am able to plot them in topography plot. Kindly let me …
-
### Feature request
Adding generation configurations to the parameters that can be tuned in a `Trainer`.
### Motivation
When defining the Optuna hyper-parameter space, I would like to invest…
-
Dear Rick and Team,
I would like to ask for help regarding how the scores for each feature are calculated from the generateCompounds() function. My main objective is to classify the features identi…
-
Hi and thanks for the work, autoGO is a very convenient package that we use a lot.
I want to ask regarding the universe/background that is used in the statistical tests of these analyses. It's the…
-
- [x] See the effect on 'method' parameter on LRP maps.
- [ ] See effect of different 'beta' parameter in the 'alpha-beta method'. Noticed differences in the online [demo](https://lrpserver.hhi.f…
-
I'm working on a CLI with multiple, dynamic sources of defaults and pretty complicated rules for what parameters are required when, so I was pretty excited to discover Cloup's constraints. However, I …
-
### Expected behavior
I'm trying out the GPSampler to hyperopt my ML pipeline, which is basically a non-deterministic, multivariate objective with real, integer and categorical parameters, and a co…
-
The conversion has a problem the matrix is not coherent
Baseline:
Software version: syngo MR XA60
Protocol name: Initialized by sequence
Detected Numaris/X version: 60
Disabling parsing of sy…
-
Q - Basic convolution without padding reduces the resolution. Is the reason feature resolution isn't reduced here because of padding?
![image](https://github.com/user-attachments/assets/dbe94191-4886…