Converting multivariate dataframe to univariate time series using R -


these head of data set using.

 hour count   <chr> <int> 1    00 22462 2    01 13293 3    02 10595 4    03  9371 5    04 14325 6    05 38598 

to perform forecasting using auto arima should convert data univariate.

make.univ(rsms,sms.hour,tname="time1", outname="multdv") 

i used above code convert univariate, gives error.

error in data.frame(timedat = rep(0:(nrepobs - 1), nrow(x)), outdat = as.vector(t(dvs))) : arguments imply differing number of rows: 4123938, 48

original dataset:

rsms [2million records(1 day data)] sample dataset : sms.hour[24 records (class(sms.hour) = "tbl_df"     "tbl"        "data.frame"] 

can me solve this?

your data univariate. i'm not sure, think lines below make data output of make.univ() call:

rsms$hour <- as.numeric(rsms$hour) names(rsms) <- c("time", "multdv") 

Comments

Popular posts from this blog

php - Vagrant up error - Uncaught Reflection Exception: Class DOMDocument does not exist -

vue.js - Create hooks for automated testing -

Add new key value to json node in java -