FrankWolfe.jl

FrankWolfe v0.2.6

Diff since v0.2.5

Merged pull requests:

• Update primal value before the callback. (#352) (@dhendryc)

FrankWolfe v0.2.5

Diff since v0.2.4

Closed issues:

• ActiveSet functions without broadcast (#343)

Merged pull requests:

• callback ordering (#350) (@matbesancon)
• up doc (#351) (@matbesancon)

FrankWolfe v0.2.4

Diff since v0.2.3

Merged pull requests:

• Quick fix in active_set.jl. (#349) (@dhendryc)

FrankWolfe v0.2.3

Diff since v0.2.2

Merged pull requests:

• ActiveSet operations required by the iterate and vertex types (#348) (@matbesancon)

FrankWolfe v0.2.2

Diff since v0.2.1

Closed issues:

• Some examples broken (#344)

Merged pull requests:

• renaming (#345) (@matbesancon)
• parenthesis (#346) (@matbesancon)
• different footer (#347) (@matbesancon)

FrankWolfe v0.2.1

Diff since v0.2.0

Merged pull requests:

• reset printing to previous (#342) (@matbesancon)

FrankWolfe v0.2.0

Diff since v0.1.22

Closed issues:

• Deal with numerically unstable Lipschitz estimates in adaptive LS (#337)

Merged pull requests:

• Breaking 0.2 (#295) (@matbesancon)
• Simplified callback state (#321) (@j-geuter)
• Fix dualstep check breaking (#330) (@matbesancon)
• fix print (#332) (@matbesancon)
• start counting iterations with 1 (#333) (@hannahtro)
• Minor change of iteration counter placement (#334) (@dhendryc)
• Discarded vertex storage (#335) (@matbesancon)
• updated counter of print_header call to 1 (#336) (@dviladrich95)
• Breaking02md (#338) (@j-geuter)
• improved and debug (#339) (@matbesancon)
• formatting (#340) (@matbesancon)
• remove PP from printing by default (#341) (@matbesancon)

FrankWolfe v0.1.22

Diff since v0.1.21

Merged pull requests:

• fix tol with epsilon (#329) (@matbesancon)

FrankWolfe v0.1.21

Diff since v0.1.20

Closed issues:

• Replace broadcast assignment with LinearAlgebra.mul! (#269)
• Set gamma_max correctly in (#317)

Merged pull requests:

• Callback-stop-unified (#313) (@dviladrich95)
• Readd gradient (#314) (@matbesancon)
• Fix signature (#316) (@matbesancon)
• Fix gamma (#318) (@dhendryc)
• up project (#319) (@matbesancon)
• Movielens bug fix (#320) (@dviladrich95)
• fix (#322) (@matbesancon)
• fix (#323) (@matbesancon)
• Added docstrings for linesearch methods (#324) (@zevwoodstock)
• Added page for line search methods (#325) (@zevwoodstock)
• adapted doc improvement to breaking-02 (#326) (@matbesancon)
• Quick fix (#327) (@dhendryc)
• fix print (#328) (@matbesancon)

FrankWolfe v0.1.20

Diff since v0.1.19

Closed issues:

• Just use the user-passed x0 to compute actual first point (#147)
• Choose which functions to include in precompilation (#220)

Merged pull requests:

• Step counter 02 (#301) (@dviladrich95)
• use function that calls @memory_mode (#302) (@hannahtro)
• Documentation revamp (#306) (@gdalle)
• Fix math in README (#307) (@gdalle)
• Update basics.md (#308) (@gdalle)
• Remove remaining math in README.md (#309) (@gdalle)
• Update README.md (#310) (@gdalle)
• up MOI to 1 (#311) (@matbesancon)

FrankWolfe v0.1.19

Diff since v0.1.18

Merged pull requests:

• RO multiplication (#303) (@matbesancon)

FrankWolfe v0.1.18

Diff since v0.1.17

Closed issues:

• Optional in-place argument for compute_extreme_point (#248)
• Incorrect Results (#298)

Merged pull requests:

• [WIP] use modify to change objective (#291) (@hannahtro)
• fix doc active set (#294) (@matbesancon)
• Fix doc (#296) (@matbesancon)
• Updated documentation for ball LMOs (#297) (@ZWoodstock)
• fix inplace scaled (#299) (@matbesancon)

FrankWolfe v0.1.17

Diff since v0.1.16

Merged pull requests:

• faster Rank-One dot (#290) (@matbesancon)
• fix dot for complex entries (#292) (@matbesancon)
• Rank one dot (#293) (@matbesancon)

FrankWolfe v0.1.16

Diff since v0.1.15

Merged pull requests:

• callback gradient homogeneous (#289) (@matbesancon)

FrankWolfe v0.1.15

Diff since v0.1.14

Merged pull requests:

• In place v (#263) (@j-geuter)
• Fix plot (#286) (@matbesancon)
• [WIP] Visualization example (#287) (@hannahtro)
• add gradient to state (#288) (@matbesancon)

FrankWolfe v0.1.14

Diff since v0.1.13

Merged pull requests:

• [WIP] Speed inference (#245) (@matbesancon)
• fix header (#278) (@matbesancon)
• [WIP] type stable eigenvalues in minimize_over_convex_hull (#279) (@hannahtro)
• Contributing (#280) (@matbesancon)
• update precompilation (#281) (@hannahtro)
• trajectory names (#282) (@matbesancon)
• separate doc action (#283) (@matbesancon)
• update callback arguments (#284) (@hannahtro)
• reduce readme (#285) (@matbesancon)

FrankWolfe v0.1.13

Diff since v0.1.12

Closed issues:

• Double64 (#273)

Merged pull requests:

• Step size strategy refactor (#259) (@matbesancon)
• Replace type of x in ActiveSet (#275) (@hannahtro)
• Removing densel1 (#276) (@matbesancon)
• memory_mode for type stability (#277) (@hannahtro)

FrankWolfe v0.1.12

Diff since v0.1.11

Merged pull requests:

• up breaking-02 (#265) (@matbesancon)
• type stability of compute_extreme_point (#266) (@hannahtro)
• test on 1.6 and stable (#267) (@matbesancon)
• Copy MathOptLMO (#270) (@dhendryc)
• Moi up (#271) (@matbesancon)

FrankWolfe v0.1.11

Diff since v0.1.10

Merged pull requests:

• tolerance keyword (#255) (@matbesancon)
• Clean up active_set.jl (#258) (@dhendryc)
• removed unnecessarily twice computed gap (#264) (@pokutta)

FrankWolfe v0.1.10

Diff since v0.1.9

Merged pull requests:

• Base.copy for active set (#249) (@matbesancon)
• Spectrahedra sets (#250) (@matbesancon)
• Updated away_frank_wolfe (#251) (@dhendryc)
• Balls examples (#252) (@j-geuter)
• Blended Pairwise Conditional Gradients (#253) (@matbesancon)
• fix scaled shifted L1 LMO (#254) (@matbesancon)