[Do Not Merge] Support divergent color ramps

[nakul-py]

Description

Addresses #881

• Currently, Color Ramp Type only supports Divergent colors such as balance, delta , curl, diff and tarn so other colormaps are still show type unknown. Enable users to set a min and max value and also get self-populated min max values for the colormap.
• Also updates Color Ramp when user setted to reverse.

Checklist

• PR has a descriptive title and content.
• PR description contains references to any issues the PR resolves, e.g. Resolves #XXX.
• PR has one of the labels: documentation, bug, enhancement, feature, maintenance
• Checks are passing.
  Failing lint checks can be resolved with:
  • pre-commit run --all-files
  • jlpm run lint

---

📚 Documentation preview: https://jupytergis--912.org.readthedocs.build/en/912/
💡 JupyterLite preview: https://jupytergis--912.org.readthedocs.build/en/912/lite

[github-actions]

👈 Launch a Binder on branch nakul-py/jupytergis/divergent

[github-actions]

Integration tests report: appsharing.space

[nakul-py]

cmocean's "balance" is cool :)

[mfisher87]

Hey Nakul! This looks really good so far! I have some thoughts on the implementation details, most importantly around saving the min & max into the project file as symbologyState.min, symbologyState.max and updating singleband pseudocolor (for raster data) to share the same behavior. We can definitely share a common component!

[mfisher87]

Hey @nakul-py I'm a bit overwhelmed with interviewing folks for an internship program. I may be a bit delayed!

[mfisher87]

Nakul and I worked on this together for a little bit today. Just leaving some thoughts we talked about -- I'll get to a more comprehensive review on Monday!

[mfisher87]

Looks like we have a bug where changing the band number in the singleband symbology menu doesn't change the rendered output. The identify tool indicates that we should see a difference between the bands!

[mfisher87]

Suggested change

	// 2. Try smallest overview if available
	// 2. Try smallest overview if available
	// A consequence of this method is that we will almost never find the **true** minimum and maximum for the whole raster, as overviews are aggregate data, averaging groups of grid cells to calculate lower-resolution grid cells. However, calculating the minimum and maximum from the entire dataset may be prohibitively expensive on CPU and memory!

[mfisher87]

Since this is for display only, classifyFunc doesn't need to receive it as an argument anymore. Let's remove it!

Suggested change

criticalValue?: number,

[mfisher87]

Suggested change

	const scaledCritical =
	// Get the critical value (divergence point in a divergent colormap) as a data value using equal interval interpolation.
	// This is only used if equal interval classification mode is selected.
	const criticalDataValue =

[nakul-py]

I removed the code related to critical value because as we removed critical value also from classification function. So not the const criticalDataValue = used nowhere.

[mfisher87]

dataMin and dataMax aren't being saved to the symbology state (we can see that the JSON isn't being updated). This is because we need to call a function to update the shared model; if we only assign to it, it won't update the model. I don't have that function name at hand and have to jump on to a meeting. Let me know if you have trouble finding it!

Though in my testing, things seem to be working well, so I wonder if this whole effect can be deleted and things will still work?

[nakul-py]

Sure we can delete specific useEffet.

[mfisher87]

Nakul and I are on a call together thinking about how we can break this PR in to smaller pieces to make it more reviewable and make it easier to add divergent colormaps in the future. Ideas:

1. Refactor useGetBandInfo into multi- and single- band versions. This will enhance performance by avoiding analyzing all bands when it's not required. See: 79a7885
2. Refactor reverseRamp to be part of ColorRampControls instead of repeated in multiple components (Graduated, Categorized). Make the color ramp selector display show as reversed when checked. Move the checkbox over to the right so that it's immediately underneath the colormap selector dropdown.
3. Extract minor type improvements, e.g. using React.Dispatch instead of callable types
4. Update the project file to store color stops in a way that they can be restored into the UI. Right now, you need to click "Classify" every time you open the symbology dialog to view the color stops. This is because the stops are stored in the project file in 0-1 space instead of data space. We can store the color stops in data space and then convert too 0-1 space before handing the information off to OpenLayers if need be. Already in progress, see: Add support for saving color stops in symbologyState #949
5. Extract color ramps to colorRamps.ts and give them types, and display the type in the dropdoown -- don't worry about divergent for now. E.g.:

export interface ISequentialColorRampDefinition
  extends IBaseColorRampDefinition {
  type: 'Sequential';
}

/* Later!
export interface IDivergentColorRampDefinition
  extends IBaseColorRampDefinition {
  type: 'Divergent';
  criticalValue: number;
} */

export interface ICyclicColorRampDefinition extends IBaseColorRampDefinition {
  type: 'Cyclic';
}
export type IColorRampDefinition =
  | ISequentialColorRampDefinition
  | IDivergentColorRampDefinition
  | ICyclicColorRampDefinition;