Forum: developers
Monitor Forum | 
Start New Thread
| thought about 'estfun' [ Reply ] By: Ott Toomet on 2016-02-15 03:20 | [forum:42947] |
|
* do we need estfun.maxim? I don't think so as "observation" is not easily defined for general maximization problems. * should we also implement 'score', identical to estfun, just not relying on sandwich? It is common to call ML gradients 'score'... I don't know, I would not do it now. * should the generic for 'gradient' sit in maxLik (like hessian) or in miscTools (like stdEr)? I don't know, would leave it as it is now. |
|
