Model the document after HiveR::HEC.

Each with it's own color and levels?

Should check for the presence of a loadings entry and exclude it or include it as appropriate.

Make a dendrogram using the pfac scores.

We need some negative peaks to make the graphical tests more comprehensive.

It used to align easily, but it seems it no longer does. I know there have been some changes in the ML toolbox it uses. Probably time to make sure the arguments are appropriate. Hopefully not a bigger problem!

We have a test data set, just need to write it.

Check for recycling etc.

This function is central and each condition checked should have a unit test. Should include the actions in .extraData. Title should read chkSpectra2D by the way.

Use gray scale for the reference spectrum. and a color for the main peaks.

Currently it does, but on the intrinsic [0...1] scale.

Doesn't seem to hurt anything but the user should be informed.

Logically goes in pfacSpectra2D. Will also need to explore how this improves sample separation.

This is probably the only way to get the color legend/scale onto the plot w/o wasting a lot of space.

plotSpectra2D needs an option to mark the regions which have been blanked out, and sumSpectra2D needs to report such regions.

This is not as clean as removeFreq. Be sure to reference the new function in chkSpectra2D.

These would be in 2D of course.

Will need to explore if this helps separations.

Need a test suite that covers all options singly and in combination.

Could pattern off ChemoSpec::plotScores3D.

Write an internal function that will do a contour plot of any 2D data set, and have plotSpectra2D and pfacLoadings and check4Gaps2D use it.

Improve the automatic choice of levels by checking the range of values and selecting an appropriate method. Also, eliminate the lowest levels of the log choice, they always seem to be in the weeds. It might be wise to create a function to do this, which could also be used by pfacLoadings.

Use pkgdown.

Function is not exported but should still be tested, perhaps using a revised/improved MUD1 data set.

See .sH for the method.

Will need to port the code from ChemoSpec. Also add to miaScores.

When importing a csv or ascii type data, for instance the existing fmt = "YXZ" we need to take into account which dimension varies the fastest.

Add to normSpectra2D.

Likely a heavily modified version of ChemoSpec::check4Gaps. Need plotting engine first.

The functions plotScree and plotScores in CSU have been modified. Need to check how the functions in this package are affected.

In particular, which graphics mode is being used/intercepted? Do we need to add .onAttach to set it silently? In principle they could set the go as in CS since the two functions will accept multiple go options.

Also, update examples if ggplot2 is used as the default mode.