LukaszDelong / GammaMDN

Gamma Mixture Density Networks
4 stars 4 forks source link

An Error in train_mixture_gamma(..."EM","ALL","NB"...) #3

Open BrotherYA opened 3 years ago

BrotherYA commented 3 years ago

new_model=train_mixture_gamma(trial=1,

  • y_data=y_data,
  • x_data=x_data,
  • no_categorical=0,
  • validation_split=0.2,
  • validation_seed=12345,
  • no_densities=3,
  • neurons=c(30,30,30),
  • neurons_alpha_prob=c(10,10,10),
  • neurons_beta=c(10,10,10),
  • neurons_categorical=0,
  • batchsize=10000,
  • learning_rate=0.002,
  • regularization_rate=0,
  • calibration_method="EM",
  • algorithm_method="All",
  • iterate_em_method="NB",
  • no_iterations=50,
  • no_epochs=25,
  • no_epochs_initial_final=50,
  • no_epochs_final=50,
  • validation_meta_model="N",
  • noise_scale=-1) fitting ... |=================================================================================================| 100% [1] "Network built" [1] "Model built" Error in model_optimize_weights[[index_update[3 + no_categorical + i]]] <- model_optimize_weights_last[[index_update[i]]] : attempt to select more than one element in integerOneIndex In addition: Warning message: In backcompat_fix_rename_lr_to_learning_rate(...) :

Error in model_optimize_weights[[index_update[3 + no_categorical + i]]] <- model_optimize_weights_last[[index_update[i]]] : attempt to select more than one element in integerOneIndex 5.eval(ei, envir) 4.eval(ei, envir) 3.withVisible(eval(ei, envir)) 2.source(paste(my_file_path_directory_codes, "gamma_mixture_calc_boosting.r", sep =

zhangc927 commented 2 years ago

I have the same error:Error in model_optimize_weights[[index_update[3 + no_categorical + i]]] <- model_optimize_weights_last[[index_update[i]]] : attempt to select more than one element in integerOneIndex