Closed robjhyndman closed 4 years ago
I cannot reproduce this issue, could you post your sessionInfo()
?
reprex::reprex(si = TRUE)
library(feasts)
#> Loading required package: fabletools
ACF(tsibble::tourism, Trips)
#> Warning in mutate_impl(.data, dots, caller_env()): Vectorizing 'cf_lag'
#> elements may not preserve their attributes
#> Warning in mutate_impl(.data, dots, caller_env()): Vectorizing 'cf_lag'
#> elements may not preserve their attributes
#> Warning in mutate_impl(.data, dots, caller_env()): Vectorizing 'cf_lag'
#> elements may not preserve their attributes
#> Warning in mutate_impl(.data, dots, caller_env()): Vectorizing 'cf_lag'
#> elements may not preserve their attributes
#> # A tsibble: 5,776 x 5 [1]
#> # Key: Region, State, Purpose [304]
#> Region State Purpose lag acf
#> <chr> <chr> <chr> <int> <dbl>
#> 1 Adelaide South Australia Business 1 0.0333
#> 2 Adelaide South Australia Business 2 0.0590
#> 3 Adelaide South Australia Business 3 0.0536
#> 4 Adelaide South Australia Business 4 0.201
#> 5 Adelaide South Australia Business 5 0.0645
#> 6 Adelaide South Australia Business 6 0.104
#> 7 Adelaide South Australia Business 7 -0.0556
#> 8 Adelaide South Australia Business 8 0.227
#> 9 Adelaide South Australia Business 9 0.0128
#> 10 Adelaide South Australia Business 10 -0.114
#> # … with 5,766 more rows
Created on 2019-08-30 by the reprex package (v0.3.0)
The lag is also not 1Q any more. Related?
Yes, the text formatting is dependent on the class which is being dropped (as stated in the warning).
Actually is the custom class cf_lag
needed? I can't see the advantage of using cf_lag
rather than difftime
as the lag variable.
It's used for a ggplot scale which has seasonality specific plot breaks. I could probably add it to the autoplot method instead, I'll have a think about it.
Closing as I am unable to reproduce this issue. Class of *CF lag index is moved to issue #82