Little explorations with HP calculators (no Prime)
|
04-01-2017, 09:45 PM
Post: #100
|
|||
|
|||
RE: Little explorations with the HP calculators
(04-01-2017 07:58 PM)pier4r Wrote: Thanks! Actually you are right, if you check the second column in the matrix it is a triangular form so I may find something to model the sum of those values. I will have to check when I have time and then I hope to post here the results. But isn't your independent variable in this case the probability (curProbV)? If that's the case, it seems that you should instead model the variance result. As an example, instead of doing lookups for the variance, you could use the following formulas. For curProbV values of 25-3825, the variance shown in the array is approximately: \[variance=5.4 \times curProbV^{0.53}\] and from 3826-7225, the variance is approximately: \[variance=7.3 \times (7250-curProbV)^{0.5}\] The resulting values aren't exactly the same as the table lookups, but they're very close. And it's not clear to me if those differences are even a problem. For example, instead of intermediate values all being "ceiled" to the next variance in the table, they will return a unique result. Is that a good or bad thing? |
|||
« Next Oldest | Next Newest »
|
User(s) browsing this thread: 2 Guest(s)