Numeric Data : How to change histogram binning (linear/log)
Keshif histograms reveal the distribution of a numeric data attribute, from its minimum to maximum. However, sometimes, there can be many small numbers and few big ones, which creates a skewed distribution, such as:
As you can see, bins of equal size (linear binning) does not produce an insightful visualization in this case. Keshif's log-scale binning avoids this problem by creating each bin for a better focused distribution, revealing trends within the packed smaller range better.
Keshif tries to pick the best mode automatically when visualizing your data.
But, you can still change it live within a second.
Changing the Histogram Scale (Linear , Log)
You can change the histogram scale using the summary configuration.
Note that the summary needs to have all positive values (no zero or negative values) to support the logarithmic scale.