create_jaatha_model.coalmodel: Use a coala model in Jaatha

Description

This creates a Jaatha model from a coala model. Simulation for this model model are conducted via the simulate function for the coala model. The parameters that are estimated must be specified via par_range and the model must not have any other named parameters. Summary statistics present in the coala model are used in Jaatha.

Usage

# S3 method for coalmodel
create_jaatha_model(
  x,
  jsfs_summary = c("sums", "folded_sums", "none", "smooth"),
  four_gamete_breaks = c(0.2, 0.5),
  mcmf_breaks = c(0.5, 0.7, 0.9),
  jsfs_part = c(1, 3),
  jsfs_part_hi = c(1, 3),
  ...,
  scaling_factor = 1,
  test = TRUE
)

Arguments

x: The coala model
jsfs_summary: The way the Joint Site Frquency Spectrum (JSFS) is further summarized. Can be sums (default), none or "smoothing". For sums, 23 different areas of the JSFS are summed up, and the sums are used as indepented Poission statistcs. For folded_sums, the same sums will be calculate from the unpolarized (folded) JSFS. This does only support two population spectra and the default partitions at the moment. For none, all entries are used as indepented Possion statistics. The value smooth is experimental so far and should not be used. This option has no effect if the JSFS is not a summary statistic of the coala model.
four_gamete_breaks: Quantiles of the real data that will be used as breaks for binning the Four Gamete test based statistic if present in the model.
mcmf_breaks: Quantiles of the real data that will be used as breaks for binning the MCMF statistic if present in the model.
jsfs_part: Partitions used for the summarizing the JSFS. This is only used if jsfs_summary is "sums". Is used as the part argument of coarsen_jsfs. Please go there for an explanation. If folded_sums is used as jsfs summary, the values of jsfs_part and jsfs_part_hi will be ignored, and their default values c(1, 3) will be used instead.
jsfs_part_hi: Same as jsfs_part, but used as part_hi argument in coarsen_jsfs.
...: Additional parameters passed on to the dispatch function.
scaling_factor: If your model is a down-scaled version of your data, you can indicated this using this value. The estimated expectation values are multiplied with this factor before the likelihood is calculated.
test: A logical indicating whether a simulation is performed to test the model.