- 13 May, 2021 2 commits
-
-
Thomas Purcell authored
Should remove any overwrite issues
-
Thomas Purcell authored
An option since while it should give a better model, it also prevents reproducibility unless OMP_NUM_THREADS=1
-
- 12 May, 2021 10 commits
-
-
Thomas Purcell authored
These tests needed to be fixed for the updated scheme
-
Thomas Purcell authored
This should fix some of the issues
-
Thomas Purcell authored
Minor changes were needed
-
Thomas Purcell authored
Some debug lines removed
-
Thomas Purcell authored
Allow for the features to find better minimum
-
Thomas Purcell authored
Remove lines used to check if the feature space is correct
-
Thomas Purcell authored
These got unstandardized at one point, now it should be back to normal
-
Thomas Purcell authored
The feature space is now the same
-
Thomas Purcell authored
-
Thomas Purcell authored
Every computational environment will have the same order of features in it
-
- 11 May, 2021 10 commits
-
-
Thomas Purcell authored
New scheme needs updated tests
-
Thomas Purcell authored
Changes necessary for new parmeterization scheme
-
Thomas Purcell authored
This is apparently too cahotic to work
-
Thomas Purcell authored
log/exp now bound internal and not external parameters, removes from_parent
-
Thomas Purcell authored
-
Thomas Purcell authored
-
Thomas Purcell authored
make it slightly more efficent
-
Thomas Purcell authored
Tests should now pass
-
Thomas Purcell authored
Increase the number of maxevals to allow for slower optimization methods
-
Thomas Purcell authored
Fix was a surpious test
-
- 10 May, 2021 8 commits
-
-
Thomas Purcell authored
For updated parameterization scheme
-
Thomas Purcell authored
Log, Exp, NegExp, and AbsDiff use it
-
Thomas Purcell authored
New subplex method took too long to find previous set up
-
Thomas Purcell authored
Gradient based methods are either too slow or chaotic
-
Thomas Purcell authored
This transfers to a scaled parameter
-
Thomas Purcell authored
Rounds a number to an arbitrary precision in base 2
-
Thomas Purcell authored
Based on the precision used to optimize them
-
Thomas Purcell authored
1) If param is fixed don't print it 2) Clean up a typo in op_node gradient
-
- 09 May, 2021 5 commits
-
-
Thomas Purcell authored
This can still be an issue even if the input feature is not constant
-
Thomas Purcell authored
new algorithm from nlopt needed lower limits
-
Thomas Purcell authored
use selected.size()
-
Thomas Purcell authored
This one seems to be more stable than LD_VAR_2 check wiht a full calc
-
Thomas Purcell authored
Use deterministic psuedo random numbers
-
- 07 May, 2021 3 commits
-
-
Thomas Purcell authored
include matlab function terms
-
Thomas Purcell authored
Used for sensitivity analysis with UQLab
-
Thomas Purcell authored
The header line now is aligned properly
-
- 30 Apr, 2021 2 commits
-
-
Thomas Purcell authored
-
Thomas Purcell authored
both base/with extra works now
-