Plot Holoviews.Image using Pandas dataframe

[HuangJunye]

I posted this questions on StackOverflow as well: https://stackoverflow.com/questions/55962945/plot-holoviews-image-using-pandas-dataframe

I'm trying to make a 2D color plot of my low temperature transport measurement. The data were taken and saved in a series of .dat files. The data looks something like this:

B (T), T(K) , V (V), I (A)

0, 4.01, -10.0, 0.001

0, 4.02, -9.9, 0.001 ...

Each file was taken at a specific magnetic field. (0 to 2 Telsa with 0.2T step) in total 11 files. I imported all the files and concatenate them into a single pandas dataframe which looks like this:

B	T	V	I
0	4.01	-10.0	0.001
0	4.02	-10.0	0.001
...	...	...	...
0.2	4.01	-10.0	0.001
0.2	4.02	-10.0	0.001
...	...	...	...
0.4	4.01	-10.0	0.001
0.4	4.02	-10.0	0.001
...	...	...	...

Now I want to make a 2D color plot with V and B as the x and y axis and I as z (color). How do I do that with Holoview.Image?

I know it can be done using using seaborn heatmap but I want to use holoview for interactive plot and other functionalities. It also works with Holoviews.HeatMap, but HeatMap has a few problems. HeatMap axes are categorical not numerical, which makes it difficult and non-intuitive to set xticks, xlim, etc.

Holoviews.Image is more suitable for my data, but it accepts only gridded data, not columnar data like pandas dataframe.

I guess, in a nutshell, my question is how to convert pandas dataframe to gridded data in order to plot in Holoviews.Image element.

[ahuang11]

One way~

import pandas as pd
import holoviews as hv
import hvplot.xarray

df = pd.DataFrame({
    'B': [0, 1, 2],
    'T': [2, 3, 4],
    'I': [5, 6, 7]
}).set_index(['B', 'T'])
df.to_xarray().hvplot.image('B', 'T', cmap='RdBu_r')

[HuangJunye]

@ahuang11 It works perfectly! Thanks a lot, Andrew!

[github-actions]

This issue has been automatically locked since there has not been any recent activity after it was closed. Please open a new issue for related bugs.