You could get past the error line
if you'd change it to something like this:
y <- ts(Train$sumq, frequency=7)
but it couldn't help further down the script.
Just checking if your issue is now resolved?
If you have time could you please take this quick survey on your Tableau Community Forums experience?
[Program Manager | Tableau Community Forums]
I believe Yuriy is correct here, when I fix the Rscript to use the variable name you created in your dataframe, sumq, it ultimately results in the library being unable to create a suitable ARIMA model for the data. One thing you can try is using Tableau's native forecasting tool, which works when you set the date to a continuous day level of detail.
Tableau's forecast is able to extract a trend, but there does not appear to be any modelable seasonal patterns in the data.