first attempt incorperatro refactor_ui changes into _core

change MSPlot to dataclass and made appropriate changes

set up reliance on config (SpectrumConfig, ChromatogramConfig...)

get rid of arguments and kwargs, instead store in the class.
Makes it less confusing as to what need to call what with now for the
most part just need to call things with "self"

this makes the interface more readable when we are combining multiple
plots into a single figure. E.g. marginal plots e.g. mirror spectrum
plot

figure or axes is a class variable (dependent on backend)
plot() method does not return anything --> legends directly created from
plot

to clean up what has to be passed allow for all variables to be part of
config except for the dataframe because that is what it is called on.

working for bokeh chromatogram currrently

rename figure to canvas for consistency across backends
for bokeh since canvas is a figure object make self.fig and self.canvas
linked

having difficulties currently with tooltips

change marker size

bug relative intensity was not being applied, update snapshot. also
update axis labels

found bug in the old snapshot with coloring that fixed

spectrum tests are stange so re-design them

new snapshots look reasonable so update them

update chromatogram snapshots because now it uses the linewidth argument

plotly tooltips index does not seem reproducible with the "by" column so remove this from the
chromatogram plot and update snapshots.

remove index from tooltips

manual looking and they look fine

note: had an error where when using a binning of "auto" and "sturges" or
binning of True (e.g. whenever use the pd.cut method) it seems that the
inveral object is never really gotten rid of properly and this causes an
error in plotly graphing. Could not figure out the error so switched to
using numpy

seems that previous updates were not saved? or just very minor
differences

for spectrum strangely the files produced were .json instead of .html
and this is because the plotly backend was used by mistake. Fix this by
fixing the conftest.py bug

snapshots verified manually and look good

seems that bokeh snapshots currently do not get checked correctly. Will
have to look into this.

however bokeh snapshots seem to be working as expected

update doc to reflect new config