mdancho84
commented
3 years ago Error Cause
The error occurs here because it's looking for missing regressors. In this case, it wants a column in your new_data data frame called "id", since this was provided during training.
> calib %>%
+ modeltime_forecast()
Using '.calibration_data' to forecast.
Error: Problem occurred during prediction. Most likely cause is missing external regressors. Try using 'new_data' and supply a dataset containing all required columns. Error in model.frame.default(mod_terms, new_data, na.action = na.action, : object is not a matrix
Error: Can't subset columns that don't exist.
x Column `.value` doesn't exist.
Run `rlang::last_error()` to see where the error occurred.Used during training:
value ~ . converts to value ~ date + id.
> training(m750_splits)
# A tibble: 282 x 3
id date value
<fct> <date> <dbl>
1 M750 1990-01-01 6370
2 M750 1990-02-01 6430
3 M750 1990-03-01 6520
4 M750 1990-04-01 6580
5 M750 1990-05-01 6620
6 M750 1990-06-01 6690
7 M750 1990-07-01 6000
8 M750 1990-08-01 5450
9 M750 1990-09-01 6480
10 M750 1990-10-01 6820
# … with 272 more rowsSolution
You'll need to provide the testing data again in modeltime_forecast().
> calib %>%
+ modeltime_forecast(new_data = testing(m750_splits))
# A tibble: 24 x 7
.model_id .model_desc .key .index .value .conf_lo .conf_hi
<int> <chr> <fct> <date> <dbl> <dbl> <dbl>
1 1 DEEPAR prediction 2013-07-01 9283. 8030. 10536.
2 1 DEEPAR prediction 2013-08-01 9234. 7981. 10487.
3 1 DEEPAR prediction 2013-09-01 9746. 8493. 10999.
4 1 DEEPAR prediction 2013-10-01 10324. 9071. 11577.
5 1 DEEPAR prediction 2013-11-01 10637. 9384. 11890.
6 1 DEEPAR prediction 2013-12-01 10446. 9193. 11699.
7 1 DEEPAR prediction 2014-01-01 10376. 9123. 11630.
8 1 DEEPAR prediction 2014-02-01 10531. 9278. 11784.
9 1 DEEPAR prediction 2014-03-01 10622. 9369. 11875.
10 1 DEEPAR prediction 2014-04-01 10575. 9322. 11828.
# … with 14 more rows
First, congratulation for modeltime, what a fantastic ecosystem you are building.
So, consider this example:
calib %>% modeltime_forecast(new_data = NULL, actual_data = m750)throws this error:Thank you Matt