1 Reply Latest reply on Aug 22, 2016 8:10 AM by Jonathan Drummey

    Prediction with logistic regression

    John Coven

      Hello all,

       

      Attached is a sample workbook using Titanic survivors' data. My goal is to (a) develop a logistic regression model of passenger survival, and (b) to use that model to predict survival for a hypothesized passenger specified by the user with parameters. I get the model to run fine (it replicates perfectly in R), but bafflingly, I cannot get it to produce a single predicted probability correctly: it either produces N copies of the correct predicted probability (if Aggregate Measures is turned off), or the mean probability of survival for the training data (if turned on). Can anyone advise how to get R in Tableau to output only a single, correct probability value?

       

      Thanks.

       

       

      - JC