Reminiscent of this question: Forecasting Time Series Groups with tslm() & tidyverse except I want to use Matt Dancho’s code at https://cran.rstudio.com/web/packages/sweep/vignettes/SW01_Forecasting_Time_Series_Groups.html
How can I use forecast::tslm() to produce grouped time series predictions from a nested data frame? The solution Rob Hyndman kindly provided used the tsibble and fable packages without nesting.
The map() of tslm() at the bottom of this code generates:
Caused by error in `formula.default()`:
! invalid formula
starter_time <- Sys.time()
library(dplyr)
library(timetk)
library(tidyr)
library(purrr)
library(lubridate)
library(forecast)
library(broom)
library(sweep)
library(zoo)
monthly_qty_by_cat2 <- bike_sales %>%
mutate(order.month = as_date(as.yearmon(order.date))) %>%
group_by(category.secondary, order.month) %>%
summarise(total.qty = sum(quantity))
monthly_qty_by_cat2_nest <- monthly_qty_by_cat2 %>%
group_by(category.secondary) %>%
nest()
monthly_qty_by_cat2_ts <- monthly_qty_by_cat2_nest %>%
mutate(data.ts = map(.x = data,
.f = tk_ts,
select = -order.month,
start = 2011,
freq = 12))
## invalid formula ERROR
monthly_qty_by_cat2_fit <- monthly_qty_by_cat2_ts %>%
mutate(fit.ets = map(data.ts, tslm, total.qty ~ trend))
