error in classification

[lusimac]

Dear colleagues

I'm trying to classify a modis tile (using deeplearning models) and it presents this error that starts with information about problems with the node during the process..

|= | 1%An error has occurred in a node and a recovery will be attempted at the end of the process
An error has occurred in a node and a recovery will be attempted at the end of the process
An error has occurred in a node and a recovery will be attempted at the end of the process
An error has occurred in a node and a recovery will be attempted at the end of the process
An error has occurred in a node and a recovery will be attempted at the end of the process
An error has occurred in a node and a recovery will be attempted at the end of the process
An error has occurred in a node and a recovery will be attempted at the end of the process
An error has occurred in a node and a recovery will be attempted at the end of the process
An error has occurred in a node and a recovery will be attempted at the end of the process
An error has occurred in a node and a recovery will be attempted at the end of the process
|= | 2%An error has occurred in a node and a recovery will be attempted at the end of the process

and when it reaches 100% it presents this error as an attachment.

100%Error in purrr::map(): In index: 1. Caused by error in .check_remote_errors(): ! 69 nodes produced errors; first error:   In index: 1. Caused by error: ! [readValues] cannot read values Backtrace:

the script used is following
cubo <- readRDS("~/Lucrencio/2020/cubos/cubo_classif_012010.rds")

#sits_timeline(cubo)
temp_cnn_model <- readRDS ("/Lucrencio/2020/modelos/tempCNN/modis_temp_cnn_012010.rds")
cube_dir <- ("/Lucrencio/2020/cubos/")
results_dir <- ("~/Lucrencio/2020/resultados/")

sits_timeline(cubo)

##Probabilidade das classes

probs = sits_classify(
cubo,
ml_model = temp_cnn_model,
memsize = 150,
multicores = 30,
output_dir = "~/Lucrencio/2020/resultados/probs/",
version = "v1",
progress = TRUE,
verbose = TRUE
)

plot(probs)

[OldLipe]

Hi @lusimac, Thank you for reporting the error. Could you share with us the training samples and the model used so that we can try to reproduce your error?

[lusimac]

Hi, Thank you for your reply and sorry for the late reaction... my colleagues informed me that they had the same problem... please find the requested data attached. The model used was tempCNN, Tile 012010 from MOD13Q1-6 data.

Any substantial questions, I am at your disposal.

Best regards

amostras_2.csv