Currently top 10 is limited to numbers which can be a limitation compared to other software out there. The request is to enhance top 10 auto filter to allow filtering out the "top 10" most popular strings in a column. To test: 1. Put in "a" in A1-A10 2. Data -> filter -> auto filter 3. Select the down arrow on the auto filter in A1 and select "top 10" Observed: Everything is filtered out Expected: A stays because it is in the top 10 string cases in the column (it's the only string case ;) )
Hi I confirm, this would be an interesting improvement. I set status to New. In the meantime a workaround requires an additional column with eg =CODE(LOWER(A1))*-1 Regards Pierre-Yves
Can you explain that workaround a bit more - I want to add that info to the wiki :)
Hi (In reply to comment #2) > Can you explain that workaround a bit more - I want to add that info to the > wiki :) The "top 10" filter works only on values. You can insert a column in which the CODE function is used to get a numeric code for the first character in a text string. First, to ignore the difference between uppercase and lowercase, the cell is converted (LOWER function) Codes corresponding to the aphabet are growing. They are multiplied by -1 so that A is before B. This workaround is limited because it only takes into account the first letter of the cell. I made this proposal without testing but it seems that however it may give a good result (see attached example) Go further: in the example calculations in column B are masked by the use of a custom cell style. That said, the numbers are always displayed when selecting the filter so it may not be necessary Regards Pierre-Yves
Created attachment 99881 [details] Top10 example
I think Joel want to filter the most frequently used strings: "The request is to enhance top 10 auto filter to allow filtering out the "top 10" *most popular strings* in a column." not alphabetic order. You can replace Pierre-Yves formula with this one: =COUNTIF(C$2:C$69;C2)
Hi (In reply to comment #5) > I think Joel want to filter the most frequently used strings This is not my bad english... these are my glasses that need to be improved lol Thank you Gérard
Still not available in current daily master build: Version: 7.3.0.0.alpha0+ / LibreOffice Community Build ID: de7356c2e0cb099fac396808b5a86a0393b48e5f CPU threads: 8; OS: Linux 4.15; UI render: default; VCL: gtk3 Locale: en-AU (en_AU.UTF-8); UI: en-US TinderBox: Linux-rpm_deb-x86_64@86-TDF, Branch:master, Time: 2021-06-14_22:38:17 Calc: threaded I like this idea, but would suggest that names should change to clarify what they do. I think the name "Top 10" can be interpreted in many different ways, hence the confusion here when using it on non-numeric data. I propose: - "Top 10" becomes "Highest 10": the 10 highest numerical values (and therefore the 10 most recent datetime values too) - new item: "10 most frequent": regardless of data type, the 10 most frequent values in the column (the enhancement request described here) - new item: "First 10": regardless of data type, the 10 first values when the column is sorted in alphabetical (unicode) order
Still not available with 24.8 Beta 1! Aargh. (In reply to Stéphane Guillou (stragu) from comment #7) > I like this idea, but would suggest that names should change to clarify what > they do. I'm not against that, but would suggest a separate bug for a name change. > I think the name "Top 10" can be interpreted in many different > ways, hence the confusion here when using it on non-numeric data. For numeric data as well. Actually, I'm just going to file that. > - new item: "10 most frequent": regardless of data type, the 10 most > frequent values in the column (the enhancement request described here) > - new item: "First 10": regardless of data type, the 10 first values when > the column is sorted in alphabetical (unicode) order These are two completely different bugs/enhancement requests.