Notifications
Clear all
Topic starter
Hello All,
I have a question. I need to get a calculation with a certain amount of precision. Since you can't define datatypes in MEL I am stuck. Take a look at my MEL trace:
MelTrace] results>"150"
MelTrace] execute>call VAL("150")
MelTrace] results>150
MelTrace] execute>150 / 3025
MelTrace] results>0.05
When I get out my calculator and divide 150 / 3025 I get: 0.0495867768595041. In order to get the calculation I need I need the value to not get rounded by default. Has anyone else run into this and found a way around it?
Thanks,
Brad
Posted : April 13, 2016 9:54 am
I tried the same thing, MEL Trace results were similar but the field appears to be holding 6 digits past the decimal, see below. I would check the calculation all the way through and see if it is really a problem. If it is you can try appending a few zeros to each value and moving the decimal place back later in the calculation.
04/13/2016 16:08:31.288 INFO Process Id #3348 ML Thread [MelTrace] Fire trigger: val("150")/3025
04/13/2016 16:08:31.288 INFO Process Id #3348 ML Thread [MelTrace] execute>>call VAL("150")
04/13/2016 16:08:31.288 INFO Process Id #3348 ML Thread [MelTrace] results>>150
04/13/2016 16:08:31.288 INFO Process Id #3348 ML Thread [MelTrace] execute>>150 / 3025
04/13/2016 16:08:31.288 INFO Process Id #3348 ML Thread [MelTrace] results>>0.05
04/13/2016 16:08:31.288 INFO Process Id #3348 ML Thread [MelTrace] execute>>end
04/13/2016 16:08:31.288 INFO Process Id #3348 ML Thread [MelTrace] results>>0.05
04/13/2016 16:08:31.288 INFO Process Id #3348 ML Thread [MelTrace] execute>>Done with trigger
04/13/2016 16:08:31.288 INFO Process Id #3348 ML Thread [MelTrace]
04/13/2016 16:08:31.288 INFO Process Id #3348 ML Thread [MelTrace] results>"0.05"
04/13/2016 16:08:31.288 INFO Process Id #3348 ML Thread [MelTrace] execute>call EVAL("0.05")
04/13/2016 16:08:31.288 INFO Process Id #3348 ML Thread [MelTrace] >>
04/13/2016 16:08:31.288 INFO Process Id #3348 ML Thread [MelTrace] Fire trigger: 0.05
04/13/2016 16:08:31.288 INFO Process Id #3348 ML Thread [MelTrace] execute>>end
04/13/2016 16:08:31.288 INFO Process Id #3348 ML Thread [MelTrace] results>>0.05
04/13/2016 16:08:31.288 INFO Process Id #3348 ML Thread [MelTrace] execute>>Done with trigger
04/13/2016 16:08:31.288 INFO Process Id #3348 ML Thread [MelTrace]
04/13/2016 16:08:31.288 INFO Process Id #3348 ML Thread [MelTrace] results>"0.05"
04/13/2016 16:08:31.288 INFO Process Id #3348 ML Thread [MelTrace] execute>1 + 2
04/13/2016 16:08:31.288 INFO Process Id #3348 ML Thread [MelTrace] results>3
04/13/2016 16:08:31.288 INFO Process Id #3348 ML Thread [MelTrace] execute>CMDINT = 3
04/13/2016 16:08:31.288 INFO Process Id #3348 ML Thread [MelTrace] results>3
04/13/2016 16:08:31.288 INFO Process Id #3348 ML Thread [MelTrace] execute>call SIZE(Array)
04/13/2016 16:08:31.288 INFO Process Id #3348 ML Thread [MelTrace] results>3
04/13/2016 16:08:31.288 INFO Process Id #3348 ML Thread [MelTrace] execute>3 <= 3 04/13/2016 16:08:31.288 INFO Process Id #3348 ML Thread [MelTrace] results>TRUE
04/13/2016 16:08:31.288 INFO Process Id #3348 ML Thread [MelTrace] execute>for TRUE
04/13/2016 16:08:31.288 INFO Process Id #3348 ML Thread [MelTrace] results>TRUE
04/13/2016 16:08:31.288 INFO Process Id #3348 ML Thread [MelTrace] execute>Array[3]
04/13/2016 16:08:31.288 INFO Process Id #3348 ML Thread [MelTrace] results>"userok(document.hold_field)"
04/13/2016 16:08:31.288 INFO Process Id #3348 ML Thread [MelTrace] execute>call EVAL("userok(document.hold_field)")
04/13/2016 16:08:31.288 INFO Process Id #3348 ML Thread [MelTrace] >>
04/13/2016 16:08:31.288 INFO Process Id #3348 ML Thread [MelTrace] Fire trigger: userok(document.hold_field)
04/13/2016 16:08:31.288 INFO Process Id #3348 ML Thread [MelTrace] execute>>Document.hold_field
04/13/2016 16:08:31.288 INFO Process Id #3348 ML Thread [MelTrace] results>>"0.049587"
04/13/2016 16:08:31.289 INFO Process Id #3348 ML Thread [MelTrace] execute>>call USEROK("0.049587")
04/13/2016 16:08:33.167 INFO Process Id #3348 ML Thread [MelTrace] results>>FALSE
04/13/2016 16:08:33.167 INFO Process Id #3348 ML Thread [MelTrace] execute>>end
04/13/2016 16:08:33.167 INFO Process Id #3348 ML Thread [MelTrace] results>>FALSE
04/13/2016 16:08:33.167 INFO Process Id #3348 ML Thread [MelTrace] execute>>Done with trigger
Posted : April 13, 2016 10:13 am
Pass the value to a string.
document.results = str(150 / 3025) /* Results should be 0.0495867768595041 */
document.results = truncate(document.results, 6) /* For 6 digits after the decimal point 0.049586*/
Then if you want fewer digits after the decimal point, use the Truncate(number, digits)
or Round(number, digits) function.
Pretty sure this will work for you.
-James
Posted : April 13, 2016 10:31 am
How much precision do you need?
Why not simply multiply the numerator by 1000 prior to the division function? That would force 3 digits of accuracy. Just remember, you will not be seeing 0.0495 but 40.95 - but that may be a quick solution if your next steps are to have some "if" statements to give certain messages.
Posted : April 14, 2016 2:13 am
Topic starter
Thank you for your response. I wasn't able to get the needed value by passing the calculation to a string. However, when I did this it did work:
document.results = eval(150 / 3025)
The problem is I can't hard code values, I need to put variables in there. So I tried this:
local heightsqred = document.height ^ 2
local bmidivide = eval(document.weight / heightsqred)
In the example above document.height is 55, so if squared you get 3025 and document.weight is 150. When I run the code this is what I get in my trace:
[MelTrace] results> "150"
[MelTrace] execute> "150" / 3025
[MelTrace] results> 0.05
[MelTrace] execute> call EVAL(0.05)
[MelTrace] results> 0.05
[MelTrace] execute> BMIDIVIDE = 0.05
[MelTrace] results> 0.05
So it looks like the division happens before the eval() if you have variables inside of the eval().
Posted : April 14, 2016 3:45 am
Topic starter
Thank you to everyone for replying. The reason I was doing this is I was given the task of getting our BMI calculator in line with the NIH website's calculator. My initial thought after looking at MEL traces is that it was a precision issue. After looking at the javascript in the NIH page it looks like their calculator manipulates the value (up) a bit depending on conditions. I was able to recreate their functions in MEL. Thank you for all the help. It is useful though to know that sometimes Centricity may round when you aren't expecting it.
Posted : April 14, 2016 5:45 am
