A specific parameter or set of parameters can be estimated by a pipeline.
This function applies the pipeline to a synthetic epidemic with sawtooth
incidence resulting from a stepped growth rate function. The lag between
synthetic input and estimate is assessed by minimising the root mean square
error of input and estimated based on different lag offsets.
Arguments
- pipeline
a function taking an input dataset and an infectivity profile as inputs and producing an estimate as output. This is the whole parametrised pipeline including any other inputs. This can be a
purrrstyle function.- ip
the infectivity profile.
A dataframe containing the following columns:
boot (anything + default(1)) - a bootstrap identifier
probability (proportion) - the probability of new event during this period.
a0 (double) - the beginning of the time period (in days)
a1 (double) - the end of the time period (in days)
Must be grouped by: boot (exactly).
A default value is defined.
- lags
a vector with the delays to test. Defaults to -10 to +30 days
Value
a lag analysis dataframe containing the estimate type and the lag
in days that the estimate is behind the actual observation
Examples
pipeline = ~ .x %>% poisson_locfit_model() %>% rt_from_incidence(ip = .y)
lag_analysis = quantify_lag(pipeline)
quantify_lag(~ rt_epiestim(.x,ip = .y))
#> # A tibble: 1 × 2
#> # Groups: estimate [1]
#> estimate lag
#> * <chr> <int>
#> 1 rt 8