2D: It would be nice to color with two scalars using the 2D colormaps also available in the image view. An important example would be to color by t-SNE coordinates, so that when we are re-coloring an image using the t-SNE coordinates, we can see more easily where these colors come from in the embedding.
Currently there is also no possibility to access the t-SNE coordinates even for 1D colormaps, there might be some extra work needed to do that: in the dimensions only the original data dimensions are shown and dragging a t-SNE onto a view that already uses the t-SNE for the positions, does not show the Color overlay but only gives Warning, data already loaded.
3D: Example: I have HD imaging data with a registered RGB photo of the same object. It would be nice to be able to use the rgb of that photo as color for the embedded points. Currently, I can drag it onto the color overlay, it then shows the name of the dataset next to Color data, but does not seem to do anything.
2D: It would be nice to color with two scalars using the 2D colormaps also available in the image view. An important example would be to color by t-SNE coordinates, so that when we are re-coloring an image using the t-SNE coordinates, we can see more easily where these colors come from in the embedding.
Currently there is also no possibility to access the t-SNE coordinates even for 1D colormaps, there might be some extra work needed to do that: in the dimensions only the original data dimensions are shown and dragging a t-SNE onto a view that already uses the t-SNE for the positions, does not show the Color overlay but only gives Warning, data already loaded.
3D: Example: I have HD imaging data with a registered RGB photo of the same object. It would be nice to be able to use the rgb of that photo as color for the embedded points. Currently, I can drag it onto the color overlay, it then shows the name of the dataset next to Color data, but does not seem to do anything.