topic Re: computing derivative numerically at each observation in SAS/IML Software and Matrix Computations https://communities.sas.com/t5/SAS-IML-Software-and-Matrix/computing-derivative-numerically-at-each-observation/m-p/181549#M1807 <HTML><HEAD></HEAD><BODY><P>I assume you want to use finite difference approximations. Finite differences are available by using the NLPFDD function. See the article <A href="http://blogs.sas.com/content/iml/2011/10/14/hints-for-derivatives/">"Optimizing? Two hints for specifying derivatives."</A></P></BODY></HTML> Wed, 25 Feb 2015 13:50:21 GMT Rick_SAS 2015-02-25T13:50:21Z computing derivative numerically at each observation https://communities.sas.com/t5/SAS-IML-Software-and-Matrix/computing-derivative-numerically-at-each-observation/m-p/181548#M1806 <HTML><HEAD></HEAD><BODY><P>As is well known that quasi maximum likelihood estimation requires computing first derivative of the individual likelihood at each observation. Does anybody know how to implement it in SAS?d</P></BODY></HTML> Wed, 25 Feb 2015 13:37:43 GMT https://communities.sas.com/t5/SAS-IML-Software-and-Matrix/computing-derivative-numerically-at-each-observation/m-p/181548#M1806 Hanyu 2015-02-25T13:37:43Z Re: computing derivative numerically at each observation https://communities.sas.com/t5/SAS-IML-Software-and-Matrix/computing-derivative-numerically-at-each-observation/m-p/181549#M1807 <HTML><HEAD></HEAD><BODY><P>I assume you want to use finite difference approximations. Finite differences are available by using the NLPFDD function. See the article <A href="http://blogs.sas.com/content/iml/2011/10/14/hints-for-derivatives/">"Optimizing? Two hints for specifying derivatives."</A></P></BODY></HTML> Wed, 25 Feb 2015 13:50:21 GMT https://communities.sas.com/t5/SAS-IML-Software-and-Matrix/computing-derivative-numerically-at-each-observation/m-p/181549#M1807 Rick_SAS 2015-02-25T13:50:21Z Re: computing derivative numerically at each observation https://communities.sas.com/t5/SAS-IML-Software-and-Matrix/computing-derivative-numerically-at-each-observation/m-p/181550#M1808 <HTML><HEAD></HEAD><BODY><P>Thank you Dr. </P></BODY></HTML> Wed, 25 Feb 2015 14:25:06 GMT https://communities.sas.com/t5/SAS-IML-Software-and-Matrix/computing-derivative-numerically-at-each-observation/m-p/181550#M1808 Hanyu 2015-02-25T14:25:06Z Re: computing derivative numerically at each observation https://communities.sas.com/t5/SAS-IML-Software-and-Matrix/computing-derivative-numerically-at-each-observation/m-p/181551#M1809 <HTML><HEAD></HEAD><BODY><P>Dr. I have defined a module that returns a vector containing individual likelihood value at each observation. Then I want to compute the gradient using NLPFDD but SAS reports that the module must return&nbsp; a scalar.</P></BODY></HTML> Wed, 25 Feb 2015 14:31:54 GMT https://communities.sas.com/t5/SAS-IML-Software-and-Matrix/computing-derivative-numerically-at-each-observation/m-p/181551#M1809 Hanyu 2015-02-25T14:31:54Z Re: computing derivative numerically at each observation https://communities.sas.com/t5/SAS-IML-Software-and-Matrix/computing-derivative-numerically-at-each-observation/m-p/181552#M1810 <HTML><HEAD></HEAD><BODY><P>The module should return the likelihood value for a specified value of the parameters, given the data.&nbsp; The entire data set is used as input to the likelihood function, so it is not correct to say that the likelihood is evaluated "at each observation."&nbsp; Look at the article <A href="http://blogs.sas.com/content/iml/2011/10/12/maximum-likelihood-estimation-in-sasiml/">"Maximum likelihood estimation in SAS/IML</A>" and see if that helps.&nbsp; Notice that the module takes a single argument, shich is the vector of parameters fr the model.&nbsp; The data are specified by using the GLOBAL clause.</P></BODY></HTML> Wed, 25 Feb 2015 14:46:29 GMT https://communities.sas.com/t5/SAS-IML-Software-and-Matrix/computing-derivative-numerically-at-each-observation/m-p/181552#M1810 Rick_SAS 2015-02-25T14:46:29Z