Talk:Histogram

Page contents not supported in other languages.
From Wikipedia, the free encyclopedia

Criticism of histograms[edit]

The article would benefit from mentioning the many valid points of criticism of the histograms. Also there lacks some information on the uncertainty of the histogram. It would be fair to mention some alternatives to histograms (e.g. empirical cdfs, kernel density estimates) Troelspedersen80 (talk) 14:29, 13 November 2023 (UTC)[reply]

Wrong figure caption[edit]

One of the figure shows the effect of choosing different bin counts for the same data, and the labels say "Bin width = 3", "Bin width = 8" and "Bin width = 29". First, it is not 8 but 6. Second, it is not the width of the bins which are set to 3, 8 and 29 but their number. So they should say "Bins = 3", etc. Andras Vanyolos (talk) 18:16, 23 February 2024 (UTC)[reply]

Square root rule and Excel[edit]

The reference to Excel using the square root rule is quite old. Recent versions of Excel, since 2016, use Scott's rule https://support.microsoft.com/en-gb/office/create-a-histogram-85680173-064b-4024-b39d-80f17ff2f4e8#bkmk_scottrefrule. Furthermore, the reference currently provided is not to official documentation, just to some notes by an economist which says "The general rule Excel uses is equal-width intervals with the number of intervals approximately equal to the square root of the number of data points." but there is no reference given to any Excel documentation here either, nor demonstration that Excel did use this rule in 2007. The documentation to legacy versions which would be relevant does not mention what rule is used to select the number of bins https://support.microsoft.com/en-gb/office/create-a-histogram-85680173-064b-4024-b39d-80f17ff2f4e8. I propose simply deleting this section, since using the square root of the number of datapoints is unmotivated by any theoretical consideration and doesn't seem to have actually been used as claimed. WikiNukalito (talk) 20:11, 8 April 2024 (UTC)[reply]