Improve adaptive stepsize for PANOC and ZeroFPR #50
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Outdated suggestions cannot be applied.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.
PANOC and ZeroFPR require a step size
gamma <= alpha/L
foralpha < 1
and the smoothness constantL
of the smooth termf
. The backtracking procedure forgamma
was not enforcing this: this fix updates the way the quadratic model forf
is computed, and makes sure thatalpha/gamma
is used as diagonal Hessian in such quadratic model (and not1/gamma
).In fact in some edge cases, the former behavior may have led the line search for
tau
to fail, later in the algorithms iteration. This fix makes sure that a more conservativegamma
is selected in those cases, reducing the chance of line search failure.