Implemented plot_span_map function

[0Hughman0]

Description of the change

A few sentences and/or a bulleted list to describe and motivate the change:

• Added analysis section to utils folder. I would suggest this could be a place to put utilities for creating interactive plots that allow users to analyse their data.
• implemented plot_span_map. This is a tool for automatically plotting a spatial map of intensity of a hyperspectral map over a number of ranges defined by a SpanROI.

Progress of the PR

• [x] Change implemented (can be split into several points),
• [x] docstring updated (if appropriate),
• [x] update user guide (if appropriate),
• [x] added tests,
• [x] added line to CHANGELOG.md,
• [x] ready for review.

Minimal example of the bug fix or the new feature

import numpy as np
import hyperspy.api as hs

from lumispy.utils.analysis import plot_span_map

s = hs.signals.Signal1D(data=np.random.random((64, 64, 1024)), 
                        axes=[{'name': 'x', 'size': 64}, 
                              {'name': 'y', 'size': 64}, 
                              {'name': 'sig', 'size': 1024}])

plot_span_map(s, nspans=2)

[codecov-commenter]

Codecov Report

Base: 98.36% // Head: 94.94% // Decreases project coverage by -3.41% :warning:

Coverage data is based on head (e8c36f2) compared to base (a3ac072). Patch coverage: 0.00% of modified lines in pull request are covered.

Additional details and impacted files

```diff @@ Coverage Diff @@ ## main #148 +/- ## ========================================== - Coverage 98.36% 94.94% -3.42% ========================================== Files 11 12 +1 Lines 550 574 +24 ========================================== + Hits 541 545 +4 - Misses 9 29 +20 ``` | [Impacted Files](https://codecov.io/gh/LumiSpy/lumispy/pull/148?src=pr&el=tree&utm_medium=referral&utm_source=github&utm_content=comment&utm_campaign=pr+comments&utm_term=LumiSpy) | Coverage Δ | | |---|---|---| | [lumispy/utils/analysis.py](https://codecov.io/gh/LumiSpy/lumispy/pull/148/diff?src=pr&el=tree&utm_medium=referral&utm_source=github&utm_content=comment&utm_campaign=pr+comments&utm_term=LumiSpy#diff-bHVtaXNweS91dGlscy9hbmFseXNpcy5weQ==) | `0.00% <0.00%> (ø)` | | | [lumispy/signals/luminescence\_spectrum.py](https://codecov.io/gh/LumiSpy/lumispy/pull/148/diff?src=pr&el=tree&utm_medium=referral&utm_source=github&utm_content=comment&utm_campaign=pr+comments&utm_term=LumiSpy#diff-bHVtaXNweS9zaWduYWxzL2x1bWluZXNjZW5jZV9zcGVjdHJ1bS5weQ==) | `100.00% <0.00%> (+5.05%)` | :arrow_up: | Help us with your feedback. Take ten seconds to tell us [how you rate us](https://about.codecov.io/nps?utm_medium=referral&utm_source=github&utm_content=comment&utm_campaign=pr+comments&utm_term=LumiSpy). Have a feature suggestion? [Share it here.](https://app.codecov.io/gh/feedback/?utm_medium=referral&utm_source=github&utm_content=comment&utm_campaign=pr+comments&utm_term=LumiSpy)

:umbrella: View full report at Codecov.
:loudspeaker: Do you have feedback about the report comment? Let us know in this issue.

[0Hughman0]

Sorry, black somehow got carried away and 'fixed' a bunch of files that aren't mine 😬.

[jlaehne]

Sorry, still did not find the time for a detailed look at the code. But already the demo video is very cool.

As suggested by @ericpre in https://github.com/LumiSpy/lumispy/issues/139#issuecomment-1231397069 it would be quite useful for any hyperspectral signal (EDX, EELS) and not only luminescence - so we should at least discuss incorporating it upstream.

For linescans, I could imagine an equivalent tool (possibly covered by the same function) that plots intensity traces for the chosen spectral ROIs.

[0Hughman0]

Great thanks.

With some guidance I would be happy to help adapt the code to be incorporated into hyperspy.

I think knowing how best to handle the various shapes of data will be the challenge here.

[jlaehne]

@LMSC-NTappy check out this PR

[jlaehne]

With some guidance I would be happy to help adapt the code to be incorporated into hyperspy.

I still think an upstream contribution would be great. To have the functionality available more quickly, a way forward could be to merge it in LumiSpy and make a release. Once it is also available in HyperSpy, we could deprecate it on the LumiSpy side.

I think knowing how best to handle the various shapes of data will be the challenge here.

Currently you have a span-ROI, but in a 2D map one would have rectangular ROIs to select parts of the data for the filtering. Would be surely nice to have that as well. Actually, once we integrate more functionality for Streak-Camera-Images, it would be helpful to have 2D signal support for LumiSpy as well.

@0Hughman0 could you solve the merge conflicts?

[0Hughman0]

Hi,

Currently rushing to write up my PhD!

Would be happy to do this, but will have to wait until likely April.

[LMSC-NTappy]

@LMSC-NTappy check out this PR

Will do!

[jlaehne]

Any time to upstream this to HyperSpy, could possibly make it into the v2.0 release scheduled for September?

[0Hughman0]

Hiya,

Sorry for the delay, thanks for your patience!

I'll have a go at getting this going swiftly.

I'm trying to find when the release is scheduled for 2.0, but no luck? I can hopefully sort in the next few days.

I guess I should make a new PR on the hyperspy repo?

Change's I'll make:

• [ ] Add the ability to pass SpanROI not just ranges
• [ ] Move into the plot module
• [ ] Try and up the test coverage!
• [ ] Fix conflicts

[jlaehne]

Sound like a good plan

I'm trying to find when the release is scheduled for 2.0, but no luck? I can hopefully sort in the next few days. In general, the idea for now mentioned in some other PRs is to release it in the course of September (though looks more like late September already I would say).

I guess I should make a new PR on the hyperspy repo? I think that would be the best strategy, one can transfer PRs, but I think it would be cleaner to start a new one - @ericpre has more experience with that though.

[0Hughman0]

This PR has been moved to one for hyperspy:

https://github.com/hyperspy/hyperspy/pull/3224