predict no longer generates confidence intervals

[joethorley]

With the new version (0.1.0), the confidence interval of the whole curve did not result when applying the “predict” to the average product of “ssd_fit_dists”

Percent est se lcl ucl dist  
   <int> <dbl> <dbl> <dbl> <dbl> <chr> 
 1 1 0.304 NA NA NA average
 2 2 0.544 NA NA NA average
 3 3 0.780 NA NA NA average
 4 4 1.01 NA NA NA average
 5 5 1.25 NA NA NA average
 6 6 1.49 NA NA NA average

[joethorley]

This is because in version 0.1.0 the default value of the ci argument changed from TRUE to FALSE.

The solution is to set ci = TRUE

pred <- predict(dist, ci = TRUE, nboot = 10000L)

[twinstan1988]

Hi,
I still have the following error after adding ci = TRUE

Error in rlogis(n = n, location = exp(lscale), scale = exp(lshape)) :
invalid arguments

[joethorley]

See issue #96

[twinstan1988]

See issue #96

Thank you!!