tlverse/hal9001

Issues

• Still have bases with interactions when setting max_degree = 1

  #111 opened 5 months ago by SeraphinaShi
  0

• Update needed ahead of Matrix 1.6-2 release

  #109 opened 7 months ago by jaganmn
  2

• Symmetric reduce_basis

  #104 opened 2 years ago by rachaelvp
  0

• lasso fits while bounding L1 norm

  #33 opened 2 years ago by nhejazi
  1

• CRAN solaris builds fail

  #96 opened 2 years ago by nhejazi
  1

• `num_knots` default does not grow with sample size?

  #98 opened 3 years ago by benkeser
  2

• Option to avoid penalizing one-way basis functions for time in pooled hazard regression

  #81 opened 3 years ago by Larsvanderlaan
  1

• python version

  #89 opened 3 years ago by matthewvowels1
  3

• coupling lambda search with CV

  #18 opened 3 years ago by nhejazi
  1

• incorporating covariance updates

  #15 opened 3 years ago by nhejazi
  1

• custom gradient descent with momentum

  #27 opened 3 years ago by nhejazi
  3

• custom IRLS implementation

  #25 opened 3 years ago by nhejazi
  1

• making HAL faster

  #22 opened 3 years ago by nhejazi
  1

• HAL fails to return a matrix of predictions when `cv_select=FALSE`

  #84 opened 3 years ago by nhejazi
  2

• Quasibinomial family

  #75 opened 3 years ago by Lauren-EylerDang
  3

• Poisson family?

  #72 opened 4 years ago by tcovert
  4

• relaxing HAL constraints from `glmnet` defaults

  #56 opened 4 years ago by nhejazi
  0

• predict method should break here but gives a wrong answer instead...

  #32 opened 4 years ago by jlstiles
  4

• Example usage

  #69 opened 4 years ago by daviddewhurst
  2

• additive models

  #34 opened 6 years ago by benkeser
  2

• License file completion

  #65 opened 4 years ago by rrrlw
  1

• Competing work

  #70 opened 4 years ago by daviddewhurst
  3

• Paper references

  #66 opened 4 years ago by rrrlw
  1

• lambda parameter ignored by predict.hal9001

  #58 opened 4 years ago by prockenschaub
  2

• `lambda.min.ratio` ignored by `glmnet` for `family = "binomial"`

  #57 opened 4 years ago by nhejazi
  1

• Issues installing the package

  #54 opened 4 years ago by Bahamyirou
  5

• binary prediction fails when `cv_select = FALSE`

  #50 opened 5 years ago by nhejazi
  0

• should replace `ml_hal_fit` with `squashed` right?

  #28 opened 5 years ago by wilsoncai1992
  2

• fitting grouped binomial

  #49 opened 5 years ago by HughSt
  2

• predict method not working when reduce_basis

  #46 opened 5 years ago by sgruber65
  1

• degrees should not be NULL by default?

  #45 opened 5 years ago by nhejazi
  1

• Conflict with data.table package

  #42 opened 5 years ago by rfherrerac
  1

• supporting active-set convergence

  #14 opened 6 years ago by nhejazi
  3

• lassi code should be class

  #20 opened 6 years ago by jeremyrcoyle
  1

• hal needs speed

  #21 opened 7 years ago by nhejazi
  0

• Use hal9001 in superlearner() in R

  #30 opened 6 years ago by TerrySun920302
  3

• logit link for binomial outcomes

  #13 opened 6 years ago by nhejazi
  4

• Dave, my faster replacement is giving different answers, I'm afraid, Dave ...

  #9 opened 6 years ago by jlstiles
  10

• family argument in fit_hal is funky

  #10 opened 6 years ago by jlstiles
  7

• centering of sparse design matrix

  #12 opened 6 years ago by nhejazi
  2

• unit tests and sanity checks

  #4 opened 7 years ago by nhejazi
  0

• custom gradient descent in Rcpp

  #7 opened 7 years ago by nhejazi
  1

• wrapper for Super Learner classic

  #3 opened 7 years ago by nhejazi
  1