The new_data dataset extends existing stats::simulate functionality by allowing the simulation to accept a new time index for simulating beyond the sample (. seed: Random generator initialisation (handled by fablelite).times: The number of simulated series (handled by fablelite).The generate method for a fable model should accept these arguments (names chosen for consistency with tidymodels): This allows models to be used for simulating new data sets, which is especially relevant for time series as often future paths beyond the training set are simulated. Much like the tidymodels opinion toward predict, generate should not default to an archived version of the training set. # … with 302 more rows # Warning: Removed 31 rows containing missing values (geom_point). We could then interpolate these missing values using the fitted values from a linear model with a trend: olympic_running %>% model( lm = TSLM(Time ~ trend())) %>% interpolate(olympic_running) # A tsibble: 312 x 4 # Warning: Removed 31 rows containing missing values (geom_point). The winning times for the 1916, 19 Olympics are missing from the dataset due to the World Wars. Tidy time series tools should allow users to interpolate missing values using any appropriate model.įor example, the tsibbledata::olympic_running dataset contains Olympic men’s 400m track final winning times. The forecast package provides the na.interp function for interpolating time series data, which uses linear interpolation for non-seasonal data, and STL decomposition for seasonal data. Often interpolated values can be taken from model’s fitted values, and some models may support more sophisticated interpolation methods. Models that can be estimated in the presence of missing values can often be used to interpolate the unknown values.
0 kommentar(er)
