Want to Minimize the Script Node number.

Question

Hey All. 
 I have requirement to manipulate the process variables receiving in process model as a parameter. I am using 3 script task node to manipulate the PV ,as the output of 1st node is using in 2nd node and output of 2nd in 3rd node. Can I minimize that? Because I want that the process model takes very less time to execute(performance should be good). 
 See the below scenario: 
 1st Script Task Node:- Expression= cond(pv!a) Output:- pv!op1 
 2nd Script Task Node:- Expression= cond(cond(pv!op1)) Output:- pv!op2 
 3rd Script Task Node:- Expression= rule!ABC(cond(pv!op2,cond(pv!op3))) Output:- pv!op3 
 
 Thanks in advance

Ravi Roshan · Answer

Yes you can do this, however you will need at least 2 nodes. Calculate first expression in Input tab of first script task as: 
 
1st Script Task Node(Input tab):- Expression= cond(pv!a) Output:- ac!op1 
1st Script Task Node(Output Tab):- Expression= cond(cond(ac!op1)) Output:- pv!op2 
2nd Script Task Node(Output Tab):- Expression= rule!ABC(cond(pv!op2,cond(pv!op2))) Output:- pv!op3 
 
Regards, 
Ravi

mohamedt808 · Answer

You can try to achieve the same in a single Expression Rule within load and some local variables . This way it will become single node .

Mike Schmitt · Answer

If the only PV you require at the end is pv!op3 (in other words, you don't actually need to keep any values in pv!op2 or pv!op1), then mohamedt808 's idea is the best bet. In fact you can even use local variables in an output expression in a script task, without necessarily even creating a new expression rule. 
 /* For Example... */

with(
 local!op1: cond(pv!a),
 local!op2: cond(cond(local!op1)),
 local!op3: rule!ABC(cond(local!op2)),
 
 /* return output from local!op3 */
 local!op3
) Also I will say, just for the record: having 3 script tasks instead of 1, when the work being done is approximately the same amount, will not make any meaningful difference in the process runtime. I do support consolidating nodes, but don't expect any performance increase from this.

aloks0189 · Answer

I agree with @Mike and @Mohamed and adding to that, in case if you have requirement to store all three in PV then in such case return dictionary as output of your rule like:

{ 
firstpv: your data, 
secondpv: your data, 
thirdpv: your data 
} 
 
While calling this rule in process , use todatasubset(rule!myRule(....)) And store this in a process variable of type datasubset, now while parsing and storing it into separate pv's you can do following: 
 
pv!firstpv-> index(pv!datasubset.data, "firstpv", {}) 
 
Similarly for rest of the PV. 
 
Hope it will be helpful