CODEWITHSUNDEEP
r
user17144
user17144

Reputation: 438

<x> should be a numeric matric or a numeric vector

I am passing xreg (external regressors/independent variables) below to auto.arima, and I get the message xreg should be a numeric matrix or a numeric vector. I checked for class(xreg) and got "tbl_df" "tbl" "data.frame". I am just trying to lag A and B, and so I introduce A_lagged and B_lagged as the lagged versions. After I put them in as columns, I remove A and B.

Can someone tell me what could be wrong? Thank you.

xreg=Model_Dataset %>%
        ungroup() %>%
        filter(Category==Cat,Date<ForecastDate+weeks(Horizon)) %>%
        select(ExtVariables)%>% 
        mutate('A_lagged',c(249,head(`A`,-1)))%>%
        mutate('B_lagged',c(269,head(`B`,-1)))%>%
        select(-`A`,-`B`)

Upvotes: 1

Views: 509

Answers (1)

akrun
akrun

Reputation: 887168

It is , in 'A_lagged', and 'B_lagged'. It should be replaced with =

library(dplyr)
Model_Dataset %>%
    ungroup() %>%
    filter(Category==Cat,Date<ForecastDate+weeks(Horizon)) %>%
    select(ExtVariables)%>% 
    mutate('A_lagged' = c(249,head(A,-1)))%>%
    mutate('B_lagged'= c(269,head(B,-1)))%>%
    select(-A,-B)

Upvotes: 2

Related Questions