RcppTskitTestLinking

This is a demo package that shows how to link to and import from RcppTskit package, which provides R access to tskit C API.

R CMD checks:

Key files

The key files you should study are:

• DESCRIPTION

• R/RcppTskitTestLinking-package.R

• src/test.cpp

• src/Makevars.in and src/Makevars.win.in

• tools/configure.R

• configure and configure.win

• cleanup and cleanup.win

Install and test

You can install and test this package by running:

# Install from GitHub
remotes::install_github("HighlanderLab/RcppTskitTestLinking")

# Load the package
library(RcppTskitTestLinking)

# Check the help page for one implemented function
?ts_xptr_num_individuals2

Now run the example from the help page. Below we mark it up to explain what is what.

# Find an example tree sequence file in RcppTskit package
> ts_file <- system.file("examples", "test.trees", package = "RcppTskit")
> ts_file
[1] "/Users/ggorjanc/Library/R/arm64/4.5/library/RcppTskit/examples/test.trees"

# Load tree sequence into R workspace
> ts <- RcppTskit::ts_load(ts_file)

# Check summary of the tree sequence
> ts
Object of class 'TreeSequence'
$ts
         property            value
1     num_samples               16
2 sequence_length              100
3       num_trees                9
4      time_units      generations
5        min_time                0
6        max_time 6.96199333719081
7    has_metadata            FALSE

$tables
        table number has_metadata
1 provenances      2           NA
2 populations      1         TRUE
3  migrations      0        FALSE
4 individuals      8        FALSE
5       nodes     39        FALSE
6       edges     59        FALSE
7       sites     25        FALSE
8   mutations     30        FALSE

# Note that we have 8 individuals in ts

# We can obtain the same number by running
> ts$num_individuals()
[1] 8

# This demo package implemented the same function
# that is available in RcppTskit, but since we here
# focus just on C++ code, we have to work with the
# pointer object (externalptr) `ts$pointer` and
# not with the tree sequence object (TreeSequence) `ts`
> ts_xptr_num_individuals2(ts$pointer)
[1] 8

Wrap up

Hopefully this demo package shows you sufficiently well on how to leverage RcppTskit for your package.

If we develop any methods that should work on the tree sequence object (TreeSequence) or the table collection object (TableCollection), then the methods could go into RcppTskit::TreeSequence or RcppTskit::TableCollection or new packages could inherit from this class and add their methods.