Post Reply 
Little explorations with HP calculators (no Prime)
04-02-2017, 04:01 PM
Post: #103
RE: Little explorations with the HP calculators
(04-02-2017 08:34 AM)pier4r Wrote:  as long as the difference between steps is close to 25 (say +/- 5),then it is all fine .

If there's a need for the variance result to be stepped, it can easily be rounded.

To see how much of a difference this would make, I took your latest version and made the following changes:

1) I enabled TEVAL in maintimed to get the final elapsed time
2) getVarFunc was replaced with the following:
    @ obtain a random curProbV, leave on stack
    \<-maxSumProbV RAND * 1 + IP

    @ apply variance formula based on range
      DUP 3826 <
      0.53 ^ 5.4 *
      7250 SWAP - \v/ 7.3 *

    @ round resulting variance to nearest 25
    25 / 0 RND 25 *

3) The getVarFunc profiling was disabled in matchResFunc so that true impact on performance could be seen

Executing maintimed on my 50g then completed in about 28 seconds. That's obviously a significant improvement, but I'm not certain whether computing the variance in this way has other ramifications. If it's OK, hopefully this brings the app performance back into a reasonable range.
Find all posts by this user
Quote this message in a reply
Post Reply 

Messages In This Thread
RE: Little explorations with the HP calculators - DavidM - 04-02-2017 04:01 PM

User(s) browsing this thread: 1 Guest(s)