If you’re looking to analyse your hyperspectral images (spectrum-images, image-images or n-dimensional- n-dimensional datasets), I can highly recommend hyperspy [1].
One of the brilliant ideas hyperspy incorporates is that we consider datasets to have a navigation dimension and a signal dimension (think, you measure a spectrum at each point on an image), and you can easily transpose between them. This means that you can «move around» on the image and see what the spectrum looks like, or transpose and see what the image looks like as a function of the spectrum.
In particular I think the model building, where you can fit components to your dataset, is really useful.
It works best with the Jedi LSP - pyright doesn’t support the way we added lazy loading / extensions to the base hyperspy package.
Hyperspy is great and the ability to "move around" n-dimensiobal datasets is a very powerful tool for the data visualization!
When I used it I missed two things compared to a similar superpower tool I used when I was working with multidimensional field test data in Matlab.
1. Ability to use "text dimensions", or non-uniformly spaced grid points.
2. Ability to select and filter on arbitrary expressions instead of by slice only.
The need for (2) is harder to grok (what's that going to do for a grid dataset???), but being able to apply a few arbitrary selection expressions is a superpower when analyzing messy 10+ dimensional data.
That, and the ability to add, on the fly, virtual dimensions for arbitrary expressions.
Someday, when I am ready to retire, I will take half a year to build this in python...
Thank you for the info! I recall looking at the available tools and thought that neither scratched my itch of flexible interactive filtering filtering and flexible interactive visualization. Great tools for either one, but not for both. But I will give xarray another look.
One of the brilliant ideas hyperspy incorporates is that we consider datasets to have a navigation dimension and a signal dimension (think, you measure a spectrum at each point on an image), and you can easily transpose between them. This means that you can «move around» on the image and see what the spectrum looks like, or transpose and see what the image looks like as a function of the spectrum.
In particular I think the model building, where you can fit components to your dataset, is really useful.
It works best with the Jedi LSP - pyright doesn’t support the way we added lazy loading / extensions to the base hyperspy package.
[1] https://hyperspy.org/