Forum: developers
Monitor Forum | Start New ThreadRE: Inf in data [ Reply ] By: Arne Henningsen on 2015-02-04 07:24 | [forum:41869] |
Yes, I agree that Inf should not lead to an error message. The same as NA seems plausible. Should a warning message be issued if observations are dropped due to Inf? |
Inf in data [ Reply ] By: Ott Toomet on 2015-02-04 06:25 | [forum:41868] |
Currently Inf in data is not handled at all and leads to an error through maxLik. Should we handle it somehow? The same way as NA-s? |