2 Replies Latest reply: Dec 6, 2014 6:55 AM by zdzichu towarzychu RSS

    A few bivalent nominal variables into one dimension

      Hi,

      Is it possible to work on Qlik Sense with data coded like that:

      (variables [internet; leaflet; friend; sex] in bivalent nominal scales)

      Nowy obraz.bmp

      and generate in QlikSense sheet charts like these below, cooperating each other (I mean interacting like it is beetwen charts in QS demo app's):

      wykres.bmpwykres2.bmp

      In other words, is this possible to turn the variables [internet; leaflet; friend] into one dimension [eg. info source] and take their average value as a measure, and at the same time, be able to categorize this comparision by the [sex] variable.

       

      Well, hope I'm not confusing and the question is interesting. I'm a beginer in data mining but I'm open to every solution - even the most mysterious script

      Thanks for help

      All the best

      Przemyslav

        • Re: A few bivalent nominal variables into one dimension
          Josh Good

          Hi Przemyslav,

           

          I would recommend split the table into two tables and joining them on ID.  One table, I've called People and the other I've call InfoSource.  I've also done a mapping to convert the numeric representation of gender to words.

           

          Below is the script I used and I've also attached the data and QVF.

           

          Please make this response and helpful or correct if applicable to help out other member when search for answers.

           

          -Josh

          Qlik

           

          TempData1:

          LOAD

              ID,

              Internet,

              Leaflet,

              Friend,

              Sex

          FROM [lib://Desktop/Data.xlsx]

          (ooxml, embedded labels, table is Sheet1);

           

           

           

           

          //Map Sex from numbers to Man or Woman

          SexMap:

          Mapping Load * inline [

          SexID, Sex

          1, Female

          2, Male

          ];

           

           

          People:

          LOAD

              ID,

              ApplyMap ('SexMap',Sex) as Gender

          Resident TempData1;

           

           

           

           

          TempData2:

          CrossTable([Information Source], Data, 1)

          LOAD ID,

              Internet,

              Leaflet,

              Friend

          Resident TempData1;

           

           

          Drop table TempData1;

           

           

          InfoSource:

          Load

          ID,

          [Information Source]

          Resident TempData2

          Where Data = 1;

           

           

          Drop table TempData2;