Alejandro de la Vega
Alejandro de la Vega
Typically, when one fetches the predictors for a set of runs, the default is to get the latest version of the predictors. However, if you clone an analysis, or edit...
In using neuroscout, I realized its easy to make many analyses, and lose the "interesting" ones. Adding an ability to 'star' analyses, and perhaps add a tab to see only...
We had some nice conversations with @soichih about deploying neuroscout-cli to brainlife.io. These are the main issue I remember discussing 1) Central authentication - Brainlife.io support auth across many OAuth2...
this would be a reasonable thing to do especially if we can track changes w/ datalad
There's a are some transformations implemented by pybids we might want to add to the frontend: - Replace - ToDense - Split - Factor - Rename - Group - Filter...
In datasets where there is more than one stimuli per run, the logic used to create `ComplexTextStimuli` in tokenized extraction won't work. When ingesting transcripts, each complete transcript (corresponding to...
For `SpectralCentroidExtractor` it would be much better to default to a hop size about an order of magnitude bigger. Currently it is around 0.02, while STFT is 0.1. @koudyk is...
Currently, all speech has been ingested as a series of individual words with onset and duration, with all punctuation and sentence structure stripped. The words are ingested as stimuli and...
Flask-JWT is no longer maintained. This is going to cause problems (and potentially is already incompatible with flask > 0.12. Alternatives include: https://github.com/vimalloc/flask-jwt-extended and https://flask-jwt-simple.readthedocs.io/en/latest/
@ccraddock made the suggestion to include eye tracking information as a regressor in neuroscout. HBN and forrest have it. The idea would be to be able to compute attention weights...