Opened 15 years ago
Closed 15 years ago
#759 closed enhancement (fixed)
An input missing cannot act as output missing in BSR
Reported by: | Víctor de Buen Remiro | Owned by: | Víctor de Buen Remiro |
---|---|---|---|
Priority: | highest | Milestone: | Numerical methods |
Component: | Math | Version: | 2.0.1 |
Severity: | blocker | Keywords: | |
Cc: |
Description
Now, input missing and output missing are handling in BSR as separated blocks that are simulated conditionally one from the other, so, they cannot share parameters.
However, this would be a great feature due it could allow to define observational and/or prior nodes of regression where the missing input of other nodes would be treated as missing output.
Actually the solution is very simple on the theoretical because we can simulate missing input and output all together in just one linear regression.
Change History (3)
comment:1 Changed 15 years ago by
Component: | Database → Math |
---|---|
Milestone: | → BSR Numerical methods |
Owner: | changed from Jorge to Víctor de Buen Remiro |
Status: | new → accepted |
comment:2 Changed 15 years ago by
Version: | → 2.0.1 |
---|
comment:3 Changed 15 years ago by
Resolution: | → fixed |
---|---|
Status: | accepted → closed |
In test_20 there is an example of mixed input-output missing.