Support for data that is not time series

[matt-sd-watson]

I came across your tool and am quite impressed with the API and examples!

I would like to apply something of this sort to a dash app with a UMAP scatter plot that will have a highly variable number of points (between 10,000-500,000+) each of which represents a human cell in a biological system.

However, since UMAP data is inherently unordered (the order is merely the order at which different entities in a data frame come ordered), and that the coordinates can be either positive or negative, this type of data is not directly compatible with the FigureResampler class.

I was wondering if there was any way to provide support for unordered and non times series data with similar functionality for dash apps?

[domerich]

I am the same, I have temperature data on the x axis, it is strictly monotonically increasing but it starts in the negative °C range so I get this error.

AssertionError: In order to perform time series aggregation, the data must be sorted in time; i.e., the x-data must be (non-strictly) monotonically increasing.

peak	temp_interpolated
0.976562	-4.434
0.024796	-4.434
0	-4.434
1.953125	-4.434
0.031471	-4.434
...	...
0.137329	-4.403
0	-4.402
0.047684	-4.402
0.976562	-4.402
0.124931	-4.402

The only work around seems to convert from °C to Kelvin

[JotaKas]

Any news on this enhancement? I am in a similar situation.

Thank you for the tool!

[domerich]

It seems to be working with the latest version

[matt-sd-watson]

@domerich Are you able to provide a dataset or minimal example to show this? With the latest version I still require data that are strictly monotonically increasing.

As an example, the dataset attached does not work in dash with v0.10.0

umap_coordinates.csv