Override Curve's count aggregator default to self_intersect=False

[droumis]

fixes #5975

This PR disables the self-intersect option for a Datashader count aggregator, which controls whether pixel-crossings of the same line contribute toward a pixel's count.

Why? In some cases, showing self-intersection counts (showing where a line crosses through the same pixel multiple times) is preferred, as it provides information about how the signal varies even when the variations involved are smaller than the pixels on the screen. However, in order to be more consistent with how a non-Datashader curve would appear, it was decided that a more reasonable default behavior would be to disable self-intersect for individual (non-categorical) curves.

As this PR is currently composed, not providing any aggregator argument for hvPlot will adopt the new default. One issue is that setting a count aggregator without any kwargs will revert to adopting Datashader's default (self_intersect=True), is this acceptable?

Code

import datashader as ds
import pandas as pd
import holoviews as hv
hv.extension('bokeh')
import hvplot.pandas

df = pd.read_parquet("./tseriesviz.parq")
df0 = df[df.sensor=='0']
df

width=600
colorbar=False
(df0.hvplot.line('time', 'value', rasterize=True, title='default [count]', line_width=1).opts(width=width, colorbar=colorbar) +\
df0.hvplot.line('time', 'value', rasterize=True, title='Set ds.count()', line_width=1, aggregator=ds.count()).opts(width=width, colorbar=colorbar) +\
 df0.hvplot.line('time', 'value', rasterize=True, title='Set ds.count(self-intersect=False)', line_width=1, aggregator=ds.count(self_intersect=False)).opts(width=width, colorbar=colorbar) +\
df0.hvplot.line('time', 'value', rasterize=True, title='Set ds.count(self-intersect=True)', line_width=1, aggregator=ds.count(self_intersect=True)).opts(width=width, colorbar=colorbar) +\
df.hvplot.line('time', 'value', rasterize=True, by='sensor', title='default [count_cat]', line_width=1).opts(width=width, colorbar=colorbar)).cols(2)

[codecov-commenter]

Codecov Report

All modified and coverable lines are covered by tests ✅

Comparison is base (e23c3e9) 88.62% compared to head (1043e13) 88.62%.

Additional details and impacted files

@@           Coverage Diff           @@
##             main    #6030   +/-   ##
=======================================
  Coverage   88.62%   88.62%           
=======================================
  Files         315      315           
  Lines       65636    65638    +2     
=======================================
+ Hits        58168    58170    +2     
  Misses       7468     7468           

Flag	Coverage Δ
ui-tests	23.27% <100.00%> (+<0.01%)	⬆️

Flags with carried forward coverage won't be shown. Click here to find out more.

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

[philippjfr]

As this PR is currently composed, not providing any aggregator argument for hvPlot will adopt the new default. One issue is that setting a count aggregator without any kwargs will revert to adopting Datashader's default (self_intersect=True), is this acceptable?

Yes, I think so. It's pretty rare that you'd provide a count aggregator since it's the default.

[hoxbro]

Thank you for the PR 💯

The failing test is unrelated.