how to read the epidist$prob_dist ouput?

[avallecam]

Given an example in the epidist object called covid_serialint, how can I read the covid_serialint$prob_dist output?

The first one is the meanlog, but the second one?

library(epiparameter)
library(distributional)

covid_serialint <-
  epiparameter::epidist_db(
    disease = "covid",
    epi_dist = "serial",
    author = "Nishiura",
    single_epidist = T
  )
#> Using Nishiura H, Linton N, Akhmetzhanov A (2020). "Serial interval of novel
#> coronavirus (COVID-19) infections." _International Journal of
#> Infectious Diseases_. doi:10.1016/j.ijid.2020.02.060
#> <https://doi.org/10.1016/j.ijid.2020.02.060>.. 
#> To retrieve the short citation use the 'get_citation' function

covid_serialint
#> Disease: COVID-19
#> Pathogen: SARS-CoV-2
#> Epi Distribution: serial interval
#> Study: Nishiura H, Linton N, Akhmetzhanov A (2020). “Serial interval of novel coronavirus
#> (COVID-19) infections.” _International Journal of Infectious Diseases_.
#> doi:10.1016/j.ijid.2020.02.060 <https://doi.org/10.1016/j.ijid.2020.02.060>.
#> Distribution: lnorm
#> Parameters:
#>   meanlog: 1.386
#>   sdlog: 0.568

# How to read this notation?

covid_serialint$prob_dist
#> <distribution[1]>
#> [1] lN(1.4, 0.32)

# This object is class `distribution`

class(covid_serialint$prob_dist)
#> [1] "distribution" "vctrs_vctr"   "list"

# so a reference would be:
# ?distributional::dist_lognormal()
# Is there a handy way to interpret this?

# just in case
# if we unlist the parameters, the output is reproducible
covid_serialint_parameters <- unlist(covid_serialint$prob_dist)

distributional::dist_lognormal(
  mu = covid_serialint_parameters[1],
  sigma = covid_serialint_parameters[2]
)
#> <distribution[1]>
#>            mu 
#> lN(1.4, 0.32)

^{Created on 2023-12-14 with reprex v2.0.2}

[joshwlambert]

@avallecam apologies for not getting back to you sooner on this, I had missed it.

The best approach is the use the get_parameters() function from {epiparameter}. The $prob_dist element is one of the most complex in the <epidist> object (because {distributional} is not the easiest to use when users are not familiar with vector classes, and it can be a <distcrete> object), therefore, the approach we take is to provide functions to extract data from this element rather than users extracting it directly, for example using $.

library(epiparameter)
covid_serialint <-
  epiparameter::epidist_db(
    disease = "covid",
    epi_dist = "serial",
    author = "Nishiura",
    single_epidist = T
  )
#> Using Nishiura H, Linton N, Akhmetzhanov A (2020). "Serial interval of novel
#> coronavirus (COVID-19) infections." _International Journal of
#> Infectious Diseases_. doi:10.1016/j.ijid.2020.02.060
#> <https://doi.org/10.1016/j.ijid.2020.02.060>.. 
#> To retrieve the short citation use the 'get_citation' function
get_parameters(covid_serialint)
#>   meanlog     sdlog 
#> 1.3862617 0.5679803

^{Created on 2024-02-29 with reprex v2.0.2}

See also the <epidist> print method which provides the parameters, get_parameters() is useful if you need to store these parameters in another object.

[avallecam]

yes, more clearer after our chat last week! thank you for closing this!