Support for aggregating the Y axis instead of the X axis

[MatthewCollinsNZ]

Hi Team,

We have just come across your library as we are having some major performance issues with plotly and need to resample, however the graphing that we are doing is graphs with depth on the y axis, so we I am wondering if the library could add the ability to choose which axis is aggregated, so that we would be able to leverage the code for vertical graphs with large numbers of points as well.

a snip of our graphing is below, we are looking at around 500,000 data points on screen in the snip.

Cheers

Matt

[lindzey]

I have a similar use case -- I'm trying to aggregate oceanographic data, where various measurements are plotted as a function of depth, with depth on the y-axis.

Is there any sort of work around for this? (The scientists I work with probably won't accept just rotating the depth plots!)

[jvdd]

Hi @lindzey and @MatthewCollinsNZ

Thanks for submitting this issue!

This is a fascinating use case. Am I understanding correctly that for depth logs, the correct approach is to aggregate along y: bucket by depth and retain the min/max measurement values in each bucket?

FlexViz (the successor of this library) is built for lazy, viewport-aware aggregation of large datasets. Its line downsampling is currently x-oriented, so I do not want to overstate the fit, but supporting vertical depth profiles is a use case I find very compelling. Perhaps its support for large 2D histograms and heatmaps may be useful if a density view is acceptable for this data

I would be very happy to help further. If you can share a small example dataset or describe the expected behaviour, that would help define the right solution.