v0.4.1

NonNegLeastSquares v0.4.1

Diff since v0.4.0

Merged pull requests:

• Fix tests (#37) (@timholy)
• Distributed -> Threads (#38) (@timholy)
• Fix inference problems (#39) (@timholy)
• Enhance docstring for nonneg_lsq (#40) (@timholy)

Closed issues:

• nonneg_lsq v0.6 deprecations (#7)
• pivot_cache with float32 data does not stop (#31)
• MLJ interface (#33)
• Different results shape in single process vs distributed (#34)
• Distributed vs threading (#36)

v0.4.0

NonNegLeastSquares v0.4.0

Diff since v0.3.0

Merged pull requests:

• Not perfect but worth caching somewhere (#30) (@caseykneale)
• tolerance for pivot_cache (#32) (@JeffFessler)

v0.3.0

NonNegLeastSquares v0.3.0

Diff since v0.2.0

Closed issues:

• can't "add" to julia 1.0.1 (#12)
• Coordinate Descent? (#22)

Merged pull requests:

• option to initialize passive set P (#28) (@JeffFessler)

v0.2.0

NonNegLeastSquares v0.2.0

Closed issues:

• Faster NNLS algorithm (#2)
• admm method does not always terminate (#4)
• Block principal pivoting method is missing step 5 from Algorithm 2 of Kim & Park (2011) (#9)
• Add support for more generic operators (#14)
• Info about upcoming removal of packages in the General registry (#18)

Merged pull requests:

• adding support for calling NNLS code using Gram matrices (#1) (@daniel-perry)
• Import NNLS implementation from NNLS.jl (#3) (@rdeits)
• Fix convergence in admm (#5) (@ahwillia)
• added step 5 from alg 2 of Kim & Park (2011) (#10) (@brossetti)
• Updates for v0.7 (#11) (@lstagner)
• Fixed SharedArray calls (#13) (@lstagner)
• Generalize operators (#15) (@platawiec)
• Remove SharedArray dependency (#16) (@lstagner)
• Project.toml, tests, and travis updates (#20) (@lstagner)
• add tagbot (#23) (@JeffFessler)