I am trying to plot transition probabilities between 10 different states of interest using -contour-, between time t and t+1.
The following code applies:
-twoway contour z d_count d_count_lead if cohort==50, title("Contour Plot: Probabilities") xtitle("d_count") ytitle("d_count_lead")-
Where z should take on transition probabilities. I want z to contain the following probabilities, which are visible in -tab d_count d_count_lead, row nofreq-
1) How could I map these probabilities (z) from -tab- to d_count and d_count_lead?
2) Would there be an alternative to plot predictions, rather than observed, transition probabilites? (Not sure if this question actually makes sense, since I would like to have an estimated probability for each level of the column variable (similar to the table above))
The following code applies:
-twoway contour z d_count d_count_lead if cohort==50, title("Contour Plot: Probabilities") xtitle("d_count") ytitle("d_count_lead")-
Where z should take on transition probabilities. I want z to contain the following probabilities, which are visible in -tab d_count d_count_lead, row nofreq-
1) How could I map these probabilities (z) from -tab- to d_count and d_count_lead?
2) Would there be an alternative to plot predictions, rather than observed, transition probabilites? (Not sure if this question actually makes sense, since I would like to have an estimated probability for each level of the column variable (similar to the table above))
