|
|
ML Fit Function and SaveData |
|
Message/Author |
|
KK posted on Thursday, July 19, 2007 - 10:37 am
|
|
|
Hi. Is is possible to get the value of the fit function (e.g., the ML fit function) to output directly? This can be obtained by dividing the chi-square statistic by the appropriate degrees of freedom, but I would have thought the value of F_ML (i.e., the ML fit function) would print (and ultimately output with a SAVEDATA command for simulations) if requested. Also, is there a way to increase the number of decimal places printed in the SAVEDATA command? I realize there are multiple formats the results can be output in (e.g., F2.0) but what do they mean? I cannot find their meaning in the manual and they do not (?) seem to be standard formating specifications. Thanks. |
|
|
You can obtain the fit function by asking for TECH5 in the OUTPUT command. It is not saved with the results. The format is a standard Fortran format statement. This is described in the FORMAT option of the DATA command. |
|
Nedim Yel posted on Friday, October 04, 2013 - 12:21 pm
|
|
|
Hi, I requested the TECH5 in the output command in order to get ML fit function value. Assuming the column under FUNCTION is the ML Fit function value, I have been trying to figure what the D stands for. Does it have the same meaning as E that's usually used in other programs such as excel and R? ITER STEP FUNCTION ABS CHANGE REL CHANGE STEP LENGTH DERIVATIVE 1 0 0.48954761D+00 0.000000D+00 -.121022D-01 1 0.49369038D+00 0.0041428 0.0084625 0.100000D+00 0.913050D-01 2 0.48948319D+00 -0.0042072 -0.0085219 0.106916D-01 0.952070D-05 Thank you |
|
|
D and E mean the same thing. D is double precision and E is single precision. |
|
Back to top |
|
|