incomplete Gamma function in SAS

[Barkamih]

Hi guys,

I am trying to use incomplete Gamma function model in SAS to fit a curve for dairy cows  lactaion ,but I do not know how to write this model in SAS using Proc Nlin .  Can anybody please help me about this ?

How can I specify the values of a,b and c in the model?

y(t)= a t^b e^-c*t      (Wood's model)

I look forward to hearing from you.

Ibrahim.

[Ksharp]

Calling @Rick_SAS

Here is an example. 

https://blogs.sas.com/content/iml/2015/06/08/fit-circle.html

[Rick_SAS]

It sounds like you want to estimate the parameters in the model for the "average cow."

But are there covariates in the model, such as age or treatment?

Please provide example data. For example, do your data look like this?

data cows;
input
cow time milk;
datalines;
1 1 1.1
1 2 1.2
1 3 1.3
1 4 1.1
2 1 1.3
2 2 1.3
2 3 1.5
2 4 1.4
3 1 0.9
3 2 1.2
3 3 1.4
3 4 1.3
4 1 1.0
4 2 1.1
4 3 1.5
4 4 1.3
;

proc sgplot data=cows;
series x=time y=milk / group=cow;
run;

[Rick_SAS]

The PROC NLIN documentation contains a very similar example. Here it is modified for your model, using the data that I previously posted:

proc nlin data=cows plots=fit; 
   parms A=1 B=0.1 C=0.2;
   bounds A B C > 0; 
   model milk = A * time**b * exp(-C*time);
   output out=Fit predicted=Pred;
run;

[Barkamih]

Hi Rick

My dataset looks like this 

I have Test_day_milk kg and Test date as this form of date dd/mm/yy but I had changed to Months only in order to have 12 months in X var. And milk kg in Var Y, Also How to additivity the Parameters A,B and c which based on what?

Alos, I'm looking for a code to plot this data as a lactation curve. 

my data set around 244000 records 

Herd COW_ID Test_Day_milk _kg Test_Date Months A41 2201 25 01/01/2014 1 A41 2201 45 02/04/2014 4 A41 2201 32 15/05/2014 5 A41 2201 25 13/07/2014 7 A42 2202 25 05/02/2014 2 A42 2202 54 18/06/2014 6 A42 2202 44 08/08/2014 8 A42 2202 45 20/10/2014 10 A42 2202 32 20/11/2014 11 A42 2202 33 25/12/2014 12 _

[Barkamih]

this code working and I got results 

but why 

A=1 B=0.1 C=0.2 ?

 sorry about that but I want to understand why you put 1 and 0.1 and 0.2 for parms 

thanks 

Regards 

[Rick_SAS · Accepted solution]

Glad you got the results you needed. I guess you discovered that the PLOTS=FIT option automatically creates the graph.

The PARMS statement sets initial guesses for the nonlinear optimization, so I guessed. You can put any numbers that you want there, especially if you have a better guess than I do. If you make a bad guess, the process might not converge. If you make a good guess, it will converge in only a few iterations. 

[Barkamih]

Great
thank you so much
I got the plots of my question, and thank you again for the explanation of parms.

kind Rgrds

Ibrahim