Created attachment 204720 [details] screen grab of what I described It's much too much work to create a frequency bar chart drawn from non numerical data. I (reluctantly) use Microsoft Team's polls to conduct lightweight UX user research on customer calls. For 1 of m, or n of m questions there are m possible responses. MS, in their wisdom export a CSV with the string results without an associated category index number and they don't include values with no responses. To get a frequency chart I have to use =UNIQUE() to create a separate data range showing one instance of each of the values in my responses and then manually add an index number to each unique response type Then for quicker completion I Autosort on the response data range and then manually assign each string an index drawn from the indexes on the UNIQUE data range Then I use the FREQUENCY(index column of response data range, index number of unique data range) function, which generates a new data range of frequencies of each of the responses (maybe a bug but it adds an extra row - see attachment - for some reason) THEN I can make a bar graph of that but for long responses, it should have an option to word wrap the labels (particularly when it's a bar graph and the number of response types isn't large. At the moment, the chart area truncates a long label, with a trailing elipsis (...), which is another bug I feel I believe this is a common research use case and it would be very useful to have a 'frequency chart' that can accept string OR numerical values in the data range and produce a bar chart of the frequency of each unique value in that data range, using the unique values as the labels, word wrapping the labels if necessary OR using an index and a key (user choice of which).
use ⇒ Data ⇒ Pivottable … and aggregate by »Count«
My proposal sounds like a reasonable and useful enhancement that's both intuitive and easy to use. My workaround, and yours are hacks that get you there eventually but are not intuitive, take too many steps and are not usable. I say they're not usable as a UX designer in enterprise software for 30 years. I'm under no illusion that this would get to the front of the queue anytime soon, but if we close the request because there's a hack available, it'll remain a hack forever.
(In reply to Greg from comment #2) > My proposal sounds like a reasonable and useful enhancement that's both > intuitive and easy to use. which proposal **exactly** did you talk about? > > … and yours are hacks that get you there eventually but are not > intuitive, take too many steps and are not usable. I say they're not usable > as a UX designer in enterprise software for 30 years. I'm under no illusion > that this would get to the front of the queue anytime soon, but if we close > the request because there's a hack available, it'll remain a hack forever. Oh! That's unfortunate, of course, if you're already intellectually overwhelmed by roundabout five mouse clicks!