Ability to customize bin size / bin index for x-axis in Data Drift plots

[varma53]

Hi Evidently team,

We are currently using Evidently for ML observability, specifically the Data Drift reports. For numerical columns, the data drift visualizations (e.g., chi-square–based plots / line graphs) use a binned representation, where the x-axis corresponds to bin indices created internally by Evidently.

At the moment, it appears that the binning strategy and bin size are fixed or automatically determined, and the x-axis shows these bin indices rather than user-controlled bins.

Questions:

1. Is there any existing parameter or configuration that allows users to:

   • Customize the number of bins or bin width, or

   • Control how the bin indices on the x-axis are created for data drift plots?

2. Additionally, we’ve observed that when the dataset size is small, each bin sometimes ends up containing only a single data point. In such cases, the drift plot appears to reflect the raw data values rather than an aggregated statistic (e.g., mean/median) over a batch of data points.

   • Is this the expected behavior?

   • Is there a way to enforce a minimum number of points per bin or control the aggregation used within each bin?

If these options are not currently supported, please let us know if there is a recommended workaround or if this is something planned for future releases.

Thanks in advance, and appreciate all the work on Evidently!