When chosing the width of histogram cells the use of equiprobale cells is advantageous over the use of equal-size cells as it ensures a better representation of the underlying distribution (see paragraph 6 of the paper of Jain and Chlamtac).
A histogram with $b$ equiprobale cells makes use of $(b - 1)$ quantiles. The needed quantile, we could estimate using the $P^2$-algorithm (Isssue #18).
When chosing the width of histogram cells the use of equiprobale cells is advantageous over the use of equal-size cells as it ensures a better representation of the underlying distribution (see paragraph 6 of the paper of Jain and Chlamtac).
A histogram with $b$ equiprobale cells makes use of $(b - 1)$ quantiles. The needed quantile, we could estimate using the $P^2$-algorithm (Isssue #18).