Open MichaelCurrie opened 7 years ago
Add a tab to the Django site that visualizes with d3.js the progression of the worm simulation tracks as it gets closer in parameter space to the cluster of 12,000+ worms in our database.
Notes from @aexbrown on this:
2-dimensional scatterplot speed, curvature
PCA on the angles is eigenworms
PCA (reduce to 3) PCA on the features
tennis racket principal axes hold a tennis
http://distill.pub/2016/misread-tsne/
More generally, being able to compare strains is what we want to be able to facilitate:
Relationship between strains then the averages
is my strain different from the wild-type reference?
clustering question --> given the average feature --> 300 average feature vectors
euclidean distance maholonobus distance
This would be the place where we visualize the behavioural comparison data discussed here: https://github.com/openworm/openworm/issues/229
Add a tab to the Django site that visualizes with d3.js the progression of the worm simulation tracks as it gets closer in parameter space to the cluster of 12,000+ worms in our database.
Notes from @aexbrown on this:
2-dimensional scatterplot speed, curvature
PCA on the angles is eigenworms
PCA (reduce to 3) PCA on the features
tennis racket principal axes hold a tennis
http://distill.pub/2016/misread-tsne/
More generally, being able to compare strains is what we want to be able to facilitate:
Relationship between strains then the averages
is my strain different from the wild-type reference?
clustering question --> given the average feature --> 300 average feature vectors
euclidean distance maholonobus distance