Error in py_call_impl(callable, dots$args, dots$keywords) in 3-changes.R

[ferygood]

Hello,

I am learning in chapter 3: Changes to workflow. After sourcing R scripts and make(plan), everything seems fine. However, when I use readd(best_model) to see the the output, it returns me an error as bellow:

>source("R/packages.R")
>source("R/functions.R")
>source("R/plan.R")
>print(plan)

A tibble: 7 x 3
  target      command                                                   format
  <chr>       <expr_lst>                                                <chr> 
1 churn_data  split_data(file_in("../data/customer_churn.csv"))       … <NA>  
2 churn_reci… prepare_recipe(churn_data)                              … <NA>  
3 run_relu    test_model(act1 = "relu", churn_data, churn_recipe)     … <NA>  
4 run_sigmoid test_model(act1 = "sigmoid", churn_data, churn_recipe)  … <NA>  
5 run_softmax test_model(act1 = "softmax", churn_data, churn_recipe)  … <NA>  
6 best_run    rbind(run_relu, run_sigmoid, run_softmax) %>% top_n(1, a… <NA>  
7 best_model  train_best_model(best_run, churn_recipe)                … keras 

>make(plan)
✓ All targets are already up to date.

>readd(best_model)
Error in py_call_impl(callable, dots$args, dots$keywords) : 
  AttributeError: 'str' object has no attribute 'decode'

Detailed traceback:
  File "/Users/yaochung41/Library/r-miniconda/envs/r-reticulate/lib/python3.7/site-packages/tensorflow/python/keras/saving/save.py", line 207, in load_model
    compile)
  File "/Users/yaochung41/Library/r-miniconda/envs/r-reticulate/lib/python3.7/site-packages/tensorflow/python/keras/saving/hdf5_format.py", line 182, in load_model_from_hdf5
    model_config = json_utils.decode(model_config.decode('utf-8'))

I did not find similar closed issues, so I would like to ask if you can help me with this problem. Many thanks!

Here is my sessionInfo:

>sessionInfo()
R version 4.0.5 (2021-03-31)
Platform: x86_64-apple-darwin17.0 (64-bit)
Running under: macOS Big Sur 10.16

Matrix products: default
LAPACK: /Library/Frameworks/R.framework/Versions/4.0/Resources/lib/libRlapack.dylib

locale:
[1] en_US.UTF-8/en_US.UTF-8/en_US.UTF-8/C/en_US.UTF-8/en_US.UTF-8

attached base packages:
[1] stats     graphics  grDevices utils     datasets  methods   base     

other attached packages:
 [1] yardstick_0.0.8   recipes_0.1.16    rsample_0.1.0     forcats_0.5.1    
 [5] stringr_1.4.0     dplyr_1.0.6       purrr_0.3.4       readr_1.4.0      
 [9] tidyr_1.1.3       tibble_3.1.2      ggplot2_3.3.3     tidyverse_1.3.1  
[13] keras_2.4.0       drake_7.13.2.9000

loaded via a namespace (and not attached):
 [1] fs_1.5.0           lubridate_1.7.10   filelock_1.0.2    
 [4] progress_1.2.2     httr_1.4.2         tools_4.0.5       
 [7] backports_1.2.1    utf8_1.2.1         R6_2.5.0          
[10] rpart_4.1-15       DBI_1.1.1          colorspace_2.0-1  
[13] nnet_7.3-15        withr_2.4.2        tidyselect_1.1.1  
[16] prettyunits_1.1.1  compiler_4.0.5     cli_2.5.0         
[19] rvest_1.0.0        xml2_1.3.2         scales_1.1.1      
[22] tfruns_1.5.0       digest_0.6.27      txtq_0.2.4        
[25] rmarkdown_2.8      base64enc_0.1-3    pkgconfig_2.0.3   
[28] htmltools_0.5.1.1  parallelly_1.25.0  dbplyr_2.1.1      
[31] htmlwidgets_1.5.3  rlang_0.4.11       readxl_1.3.1      
[34] rstudioapi_0.13    generics_0.1.0     jsonlite_1.7.2    
[37] tensorflow_2.4.0   magrittr_2.0.1     Matrix_1.3-2      
[40] Rcpp_1.0.6         munsell_0.5.0      fansi_0.4.2       
[43] reticulate_1.20    lifecycle_1.0.0    furrr_0.2.2       
[46] pROC_1.17.0.1      stringi_1.6.2      whisker_0.4       
[49] yaml_2.2.1         MASS_7.3-53.1      storr_1.2.5       
[52] plyr_1.8.6         grid_4.0.5         parallel_4.0.5    
[55] listenv_0.8.0      crayon_1.4.1       lattice_0.20-41   
[58] haven_2.4.1        splines_4.0.5      hms_1.1.0         
[61] zeallot_0.1.0      knitr_1.33         pillar_1.6.1      
[64] igraph_1.2.6       base64url_1.4      codetools_0.2-18  
[67] reprex_2.0.0       glue_1.4.2         evaluate_0.14     
[70] modelr_0.1.8       png_0.1-7          vctrs_0.3.8       
[73] cellranger_1.1.0   gtable_0.3.0       future_1.21.0     
[76] assertthat_0.2.1   xfun_0.23          gower_0.2.2       
[79] prodlim_2019.11.13 broom_0.7.6        class_7.3-18      
[82] survival_3.2-10    timeDate_3043.102  lava_1.6.9        
[85] globals_0.14.0     ellipsis_0.3.2     ipred_0.9-11   

[wlandau]

I suspect something is off with the Python environment, maybe Keras or TensorFlow (or maybe reticulate). You can confirm using the following reproducible example that does not use drake. If it gives you the same error (and I suspect it will) then I recommend reinstalling keras, tensorflow, and/or reticulate, possibly the GitHub development versions. You might also try reinstalling your Python environment and packages, e.g. reticulate::install_miniconda() and/or keras::install_keras(). If all else fails, https://community.rstudio.com will probably give better advice than I can, and there is an RStudio Cloud workspace at https://rstudio.cloud/project/627076.

# Run a model.
library(keras)
mnist <- dataset_mnist()
x_train <- mnist$train$x
y_train <- mnist$train$y
x_test <- mnist$test$x
y_test <- mnist$test$y
x_train <- array_reshape(x_train, c(nrow(x_train), 784))
x_test <- array_reshape(x_test, c(nrow(x_test), 784))
x_train <- x_train / 255
x_test <- x_test / 255
y_train <- to_categorical(y_train, 10)
y_test <- to_categorical(y_test, 10)
model <- keras_model_sequential() 
model %>% 
  layer_dense(units = 256, activation = 'relu', input_shape = c(784)) %>% 
  layer_dropout(rate = 0.4) %>% 
  layer_dense(units = 128, activation = 'relu') %>%
  layer_dropout(rate = 0.3) %>%
  layer_dense(units = 10, activation = 'softmax')
model %>% compile(
  loss = 'categorical_crossentropy',
  optimizer = optimizer_rmsprop(),
  metrics = c('accuracy')
)
model %>% fit(
  x_train, y_train, 
  epochs = 30, batch_size = 128, 
  validation_split = 0.2
)

# Save the model.
tmp <- tempfile()
save_model_hdf5(object = model, filepath = tmp)

# Do you get an error when you read the model?
load_model_hdf5(tmp)