feat: density heatmap

[jnumainville]

Fixes #22
This doesn't add every argument, but does add the core ones and trivial ones

Here is a good example of the basic ways this can be used, either with x, y, and histfunc="count", or x, y, z, and any histfunc, as well as other arguments to control the binning as, like histogram, how someone chooses to bin has a large impact on the visualization

import deephaven.plot.express as dx
from deephaven import new_table, agg

from deephaven import time_table

result = time_table(period="PT1S").update(["X = randomGaussian(10, 3)", "Y = randomGaussian(10, 3)", "Z = randomGaussian(10, 3)"])

density_heatmap = dx.density_heatmap(result, "X", "Y", title="Test", range_bins_x=[0, 20], range_bins_y=[0, 20], nbinsx=20, nbinsy=20)
density_heatmap_z = dx.density_heatmap(result, "X", "Y", "Z", title="Test", histfunc="std", range_bins_x=[0, 20], range_bins_y=[0, 20], nbinsx=5, nbinsy=5)

[jnumainville]

Put this up for review but going to add one more arg

      default_bin_value: The value to use for bins that have no data.
        If None and histfunc is 'count' or 'count_distinct', 0 is used.
        Otherwise, if None or 'NaN', NaN is used.

[mofojed]

Some other density_heatmap functions I tested with, adapted from https://plotly.com/python/2D-Histogram/#density-heatmaps-with-plotly-express:

import deephaven.plot.express as dx
tips = dx.data.tips()
f_tips = dx.density_heatmap(tips, x="total_bill", y="tip")
f_tips2 = dx.density_heatmap(tips, x="total_bill", y="tip", nbinsx=20, nbinsy=20, color_continuous_scale="Inferno")

iris = dx.data.iris()
f_iris = dx.density_heatmap(iris, x="petal_length", y="petal_width", z="sepal_length", histfunc="avg")