Move transformation used in GlobalBootstrap into Traits #2135
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Also, use exp instead of atan transformation for discount factors. This is an improvement on lballabio#2120 based on these observations: * Rates-based traits don't need any constraints in the optimization, so we don't need any transformation with them. This removes unnecessary computations. * Discount factors are only bounded on one side (have to be positive), so instead of estimating a max value to use atan, we can simply use exp. This is faster and uses fewer magic numbers. Another issue with atan is that it changes the gradients too much when the value is far from the middle of the range. This pushes optimization away from the correct solution when rates are low. This makes it hard to find value of maxDF that works for both very low and very high rates.
eltoder
force-pushed
the
feature/global-bootstrap-transform
branch
from
e709cb6 to
c3c27e2
Compare
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
Also, use exp instead of atan transformation for discount factors. This is an improvement on #2120 based on these observations:
Rates-based traits don't need any constraints in the optimization, so we don't need any transformation with them. This removes unnecessary computations.
Discount factors are only bounded on one side (have to be positive), so instead of estimating a max value to use atan, we can simply use exp. This is faster and uses fewer magic numbers.
Another issue with atan is that it changes the gradients too much when the value is far from the middle of the range. This pushes optimization away from the correct solution when rates are low. This makes it hard to find value of maxDF that works for both very low and very high rates.