Thanks Daniel, that's a lot clearer now.
On 2018/09/11 18:05:11, Daniel Gruno <[email protected]> wrote: > On 09/11/2018 08:03 PM, Sharan Foga wrote: > > Hi Daniel > > > > Thanks for the response. I've included some comments inline. > > > > On 2018/09/11 10:40:46, Daniel Gruno <[email protected]> wrote: > >> On 09/11/2018 12:23 PM, Sharan Foga wrote: > >>> Hi All > >>> > >>> On the mood analysis screen https://s.apache.org/YImU > >>> > >>> I’ve noticed the toggle on the mood analysis so would like to know > >>> exactly what this means. > >>> > >>> Also if you click the box to toggle between the two modes then it’s not > >>> clear from the screen when you are using ‘Relative’ mode and when you are > >>> using ‘Comparative’ mode. > >> > >> The text in the charts change when you tick the box, and should explain > >> what it is. > > > > Ah.. I see that now but am glad to hear the elaboration as it wasn't clear > > to me. > > > > If that's not clear enough, let me elaborate: > >> > >> Standard (unticked) mode > > > > So does standard mode = relative mode? > > "relative/comparative" is *one* mode, it's either off or on :) > two words for the same mode. > > If enabled, it compares moods to *all* lists it has, regardless of > filtering. > If off, it only shows how that list (or those lists) are. > > > > > shows the moods as an 'absolute' value > >> unrelated to the average across lists. Moods are still relative to each > >> other (in that the median is close to 100), but they only reflect the > >> list(s) you have chosen to analyze. > > > > Ok so to me this means that the relative mode shows the mood analysis say > > across one project if that is the filter being used. So with my Httpd > > example - in relative mode this gives me an indicator of the general mood > > profile for Httpd based on and in relation to the other moods that the > > Httpd project generates. > > > > And the mood gauge in relative mode gives me an overall indication of the > > filtered project or view (e.g Httpd) based on the moods found in the > > sources for the filter. So the more positive the moods found, the higher > > the rating will be. > > > > If my understanding is wrong then please correct me as I really need to get > > this as it is going to be part of my research :-) > > > >> > >> Relative mode (ticked) > > > > Should this be Comparative mode? > > > > uses the same values, but compares them to the > >> overall moods found in *all* lists in the database. > > > > So does this mean that you are comparing against all the sources for the > > organisation or if you have multiple organisations will it compare against > > all organisations? > > > > From what you say, it sounds like comparative mode (if that is what it is) > > compares the mood profile against everything. So Kibbletest will compare > > Httpd's mood profile against all the other projects loaded into Kibbletest. > > > > > > Thus, on the right > >> side, a score of 50% or above means you are doing as well or better than > >> other lists in terms of happy moods, and a score lower than that would > >> mean the list(s) don't quite reflect the overall mood of all lists. On > >> the left side, the values are relative to overall mood as well, in that > >> 100 means 'same as all the other lists', anything higher means a more > >> prevalent occurrence on the list(s) you picked, whereas a value lower > >> than 100 means it's less frequent there compared to the overall average. > >> > > > > OK this is good and this is clearer > > > > So now I'm after more :-) Is there a way to compare the mood profiles of > > two different projects. For example, if I wanted to compare the profiles of > > Httpd and Beam, how could I do that? > > > >> I hope that helps, and if we need to work on the UI explanations, maybe > >> add a document to the documentation, then so be it :) > > > > It does - thanks. I'd probably be happier with clearer explanations so will > > document what I can as I'm going through. > > > > Thanks > > Sharan > > > >> > >> With regards, > >> Daniel. > >> > >>> > >>> Thanks > >>> Sharan > >>> > >> > >> > >
