pyensemble/wildwood

Issues

• Feature Importance

  #117 opened a year ago by atikul-islam-sajib
  0

• import error

  #114 opened a year ago by atikul-islam-sajib
  1

• Update package for modern Sklearn (>=1.1) and other package requirements

  #113 opened a year ago by vtsanders
  10

• Raise to python 3.8

  #96 opened a year ago by stephanegaiffas
  0

• Deal with columns with a single value in Encoder

  #110 opened 3 years ago by stephanegaiffas
  0

• Remove the FutureWarning: The `inplace` parameter in pandas.Categorical.set_categories is deprecated

  #109 opened 3 years ago by stephanegaiffas
  0

• Documentation / docstring update

  #108 opened 3 years ago by stephanegaiffas
  0

• Disable all computations related to aggregation when `aggregation=False`

  #107 opened 3 years ago by stephanegaiffas
  0

• Make cache work

  #106 opened 3 years ago by stephanegaiffas
  0

• 2D Histograms instead of 3D for less memory usage

  #105 opened 3 years ago by stephanegaiffas
  0

• to handle number of modalities > 256

  #48 opened 3 years ago by yiyang-yu
  1

• High is out of bounds for int32

  #89 opened 3 years ago by alailang
  1

• Bug in

  #100 opened 3 years ago by stephanegaiffas
  1

• Put @jit flags in .utils and check that these are the ones used everywhere

  #101 opened 3 years ago by stephanegaiffas
  1

• Factories instead of higher order functions

  #77 opened 3 years ago by stephanegaiffas
  0

• Extra Encoder improvements

  #98 opened 3 years ago by stephanegaiffas
  0

• Many (normal) warnings are raised during tests

  #94 opened 3 years ago by stephanegaiffas
  0

• Encoder improvements

  #80 opened 3 years ago by stephanegaiffas
  1

• Version the poetry.lock file to avoid resolving dependencies in travis ?

  #97 opened 3 years ago by stephanegaiffas
  0

• Serialization of ForestClassifier and ForestRegressor

  #90 opened 3 years ago by stephanegaiffas
  1

• Check and test that ForestClassifier and ForestRegressor can be serialized

  #85 opened 3 years ago by stephanegaiffas
  1

• Correct some docstrings

  #84 opened 3 years ago by stephanegaiffas
  1

• Change step default for regression

  #87 opened 3 years ago by stephanegaiffas
  0

• Change and add criterion for regression

  #86 opened 3 years ago by stephanegaiffas
  0

• Bug while predict with a single node tree

  #30 opened 4 years ago by yiyang-yu
  0

• Remove all Numba function prototypes to remove eager compilation ?

  #79 opened 3 years ago by stephanegaiffas
  0

• Implement NodeContext differences

  #67 opened 4 years ago by stephanegaiffas
  0

• Activate the fastmath flag in jit at important places

  #76 opened 3 years ago by stephanegaiffas
  0

• Handle missing values

  #11 opened 4 years ago by stephanegaiffas
  0

• We can predict without binning : threshold instead of bin_threshold

  #9 opened 4 years ago by stephanegaiffas
  4

• Checks that improved splitting of categoricals indeed improves for regression and binary classification

  #35 opened 3 years ago by stephanegaiffas
  1

• Do pre-compilation works when n_jobs > 2 ?

  #25 opened 3 years ago by stephanegaiffas
  1

• Write a first version of the documentation

  #81 opened 3 years ago by stephanegaiffas
  1

• Implement entropy criterion

  #70 opened 3 years ago by stephanegaiffas
  0

• Split ordering for multiclass="multinomial"

  #47 opened 3 years ago by stephanegaiffas
  1

• Make the NUMBA_DISABLE_JIT=1 mode work again

  #17 opened 3 years ago by stephanegaiffas
  1

• anonymise every file

  #66 opened 4 years ago by yiyang-yu
  0

• Understand why this fails on Higgs

  #34 opened 4 years ago by stephanegaiffas
  2

• A figure similar to Figure 1 of AMF

  #60 opened 4 years ago by yiyang-yu
  1

• Put back `min_sample_split` parameter

  #33 opened 4 years ago by stephanegaiffas
  1

• Add properly test for categorical features

  #44 opened 4 years ago by yiyang-yu
  1

• An experiment showing that we need less trees to reach good test error

  #37 opened 4 years ago by stephanegaiffas
  0

• Experiments showing the heatmap for the step and dirichlet hyper-parameters

  #36 opened 4 years ago by stephanegaiffas
  0

• is there a problem while 'ovr' and categorical feature?

  #46 opened 4 years ago by yiyang-yu
  2

• Categorical features with < max_bins modalities

  #10 opened 4 years ago by stephanegaiffas
  0

• Dirichlet = 0.0 and aggregation

  #40 opened 4 years ago by stephanegaiffas
  0

• Continuous integration with pytest, doc and correct project.toml file

  #22 opened 4 years ago by stephanegaiffas
  1

• Add tests for statistical performance on small datasets

  #23 opened 4 years ago by stephanegaiffas
  1

• Add tests for computational speed on several datasets n large/small and d large/small

  #24 opened 4 years ago by stephanegaiffas
  0

• Add an example with weighted depths for binary classification

  #16 opened 4 years ago by stephanegaiffas
  0