) provides wonderful UI for histograms.
But now we have 1 million double numbers that need to compute histogram. How is its performance? Will it be very slow or even crash?
We can pre-process the data and store the ranges of bins with corresponding count in each bin. But it is not the data could be accept by the histogram API. It is possible to draw column charts with some hack (E.g., set the value to be the middle of range), but it doesn't seem to be a elegant solution. Do you have any suggestions?