CODEWITHSUNDEEP
rtime-seriesdeep-learningh2o
dhinar1991
dhinar1991

Reputation: 871

Error fitting the H2Oframe in R

This is my first time with deep learning using H2O Package in R. When i try to build the model, it shows error which i am unable to correct.

Here is my data

> head(d_1)
               x vibration_x    Speed
21892 1497340740       0.260 1224.601
21893 1497340800       0.100 1214.440
21894 1497340860       0.840 1218.984
21895 1497340920       0.125 1207.892
21896 1497340980       0.475 1206.744
21897 1497341040       0.025 1241.057

I am converting this dataframe to H2O type.

d_Hex_new <- as.h2o(d_1, destination_frame = "path_train")
> d_Hex_new
           x vibration_x    Speed
1 1497340740       0.260 1224.601
2 1497340800       0.100 1214.440
3 1497340860       0.840 1218.984
4 1497340920       0.125 1207.892
5 1497340980       0.475 1206.744
6 1497341040       0.025 1241.057

[8500 rows x 3 columns]

When i try to fit the model, it shows error

model <- h2o.deeplearning(x = Speed, y = vibration_x, data =  d_Hex_new,
                                                      validation = v_Hex_new,
                                                      activation = "Rectifier",
                                                      hidden = c(50, 50, 50),
                                                      epochs = 100,
                                                      classification = FALSE,
                                                      balance_classes = FALSE)

Error in h2o.deeplearning(x = Speed, y = vibration_x, data = d_Hex_new, : unused argument (data = d_Hex_new)

May i know Why it is not taking the H2Oframe data? Please help me to correct it. Thanks,D

Edited:

model <- h2o.deeplearning(x = Speed, y = vibration_x, training_frame =  d_Hex_new,
                          validation_frame = v_Hex_new,
                                                      activation = "Rectifier",
                                                      hidden = c(50, 50, 50),
                                                      epochs = 100,
                                                      balance_classes = FALSE)

Error in is.H2OFrame(x) : object 'Speed' not found

Upvotes: 0

Views: 364

Answers (1)

YCR
YCR

Reputation: 4002

There is no data parameter in the h2o.deeplearning function.

Try to replace it by the parameter training_frame.

Also, validation is actually validation_frame and classification does not exist.

And y should be the name of your variable in between double quote and x a vector containing the name or the indices of the predictors variables.

The documentation: h2o doc

Upvotes: 1

Related Questions