compare two scores or current vs previous values

mikereaves
Tera Expert

‎07-23-2025 02:19 PM

My client has a need to trend pm_project.work_effort at the record level.
I'd like to keep the solution entirely in PA rather than adding fields and/or business rules, etc.

The field: 'work_effort' is labeled: "Actual effort".
Can I use Bucket groups?

I've created a bucket group w/ three buckets
- increased (1-3)
- decreased (4-6)
- unchanged (7-9)

I've created a script: 'work effort direction'

var direction = function(curr_we, prev_we){

if (curr_we > prev_we){
    return 1;
    }
if (curr_we < prev_we){
    return 4;
    }
if (curr_we == prev_we){
    return 7;
   }
};
direction(current.work_effort, previous.work_effort);


I'd like to create an indicator "count of projects by work effort direction" with breakdown "PRJ Work Effort Direction".

The data collection job gets an error:
"Error during JavaScript evaluation com.snc.pa.dc.ScriptException: Cannot read property "work_effort" from nullTypeError: Cannot read property "work_effort" from null (<refname>; line 13) in script: var direction = function(curr_we, prev_we){...

In a PA Script, am I able to refer to "current" & "previous" ?

ANY COMMENTS GREATLY APPRECIATED@!@@@

0 Helpfuls
  • 232 Views
1 REPLY 1

wojasso
Giga Guru

‎07-23-2025 02:32 PM

Hi @mikereaves

Bucket groups in Performance Analytics are designed to classify a single numeric value into discrete ranges. They don’t automatically keep track of change over time — that’s why you’re trying to calculate an increase/decrease yourself.

In a PA collection script you can reference the current and previous sample, but be aware that previous will be null on the first record for a time series or whenever there’s no earlier sample. That’s why your code fails — you are unconditionally reading previous.work_effort.

  • Add a null check so that you handle the first sample gracefully. You can also default to the current value if there is no previous record.
  • Cast the work effort values to numbers in case they are stored as strings.
  • Return one of the bucket keys (1, 4 or 7) based on the comparison.
var curr = current.work_effort || 0;
var prev = (previous && previous.work_effort) || curr;
// Convert to floats
curr = parseFloat(curr);
prev = parseFloat(prev);
if (curr > prev) {
    return 1; // increased
}
if (curr < prev) {
    return 4; // decreased
}
return 7; // unchanged

If you use this script on your indicator, your bucket group (1‑3 = increased, 4‑6 = decreased, 7‑9 = unchanged) will evaluate correctly and the data collection job will no longer throw a null error.

You can then define your indicator as a “count of projects” grouped by the custom breakdown you created.



💥 Was this answer useful? 👉 If so, click 👍 Helpful 👍 or Accept as Solution ✅ 💡🛠️😀🙌

0 Helpfuls