DIYABC RF V1.0

Command line help

$ diyabc -h
USAGE :
-p <directory of header.txt>
-r <number of required data sets in the reftable>
-i <name given to the analysis
-c <coverage percentage for poolseq poisson law
-g <minimum number of particles simulated in a single bunch (default=100)>
-m <multithreaded version of the program
-q to merge all reftable_$j.bin 
-Q to merge all reftableRF_$j.bin 
-s <seed for the random generator (deprecated!!!)>
-t <required number of threads>
-w <computer's number if in a cluster (0 by default) >
          (each computer in the cluster is numbered between 0 and the maximal number of computers in the cluster.)
-l for producing a csv reftable file (reftable.csv) and a scenario file (scenario.txt) used by program rf
-x for translating a reftable.bin in reftable.txt
-y for translating a reftable.bin in reftable.txt writing summary statistics before parameters

-z <path/RNGfilename.bin> write the number of streams of the RNG into path/RNGfilename_cores.txt

-n for INITIALIZATION OF THE PARALLEL RNG'S (with parameters as a string including the following options separated par a semi-colon)
           t:<maximal number of the threads (per computers if cluster, 16 by default)>
           c:<maximal number of computers in the cluster (1 by default)>
           s:<seed of the first RNG (1 by default)>
           f:<forcing creation of new RNG's and overriding the old ones>

-d for ABC PRIOR/SCENARIO CHECKING (idem)
           a:<p for PCA, l for locate observed, pl for both>

-k to SIMULATE DATA FILES

-o to simulate summary statistics from a text file containing scenario and parameter values

-R <activate all stats (if empty) or selected stats families
option -p is compulsory

Prebuilt binaries for Windows/MacOS/Linux

There are avalaible at the Releases page. Please note that those are only x64 binaries.

Build

Prerequisites :

a C++17 (gcc-7 and ulterior version for example) compiler
OpenMp for vastly improved simulator performance on multicore machines
CMake

Linux & MacOS

git clone --recurse-submodules https://github.com/diyabc/diyabc.git
cd diyabc
mkdir build
cd build
cmake ../
cmake --build . --config Release

Windows

git clone --recurse-submodules https://github.com/diyabc/diyabc.git
cd diyabc
mkdir build
cd build
cmake ..\
cmake --build . --config Release

DIYABC's Projects

abcranger

ABC random forests for model choice and parameter estimation, pure C++ implementation

desktopdeployr

A framework for deploying self-contained R-based applications to the desktop

diyabc

Standalone ABC simulator for DIYABC, RF version

jcad2019

Poster for JCAD2019 presenting AbcRanger

Publication Materials for "Extending Approximate Bayesian Computation with Supervised Machine Learning to Infer Demographic History from Genetic Polymorphisms Using DIYABC Random Forest" in *Molecular Ecology Resources* special issue

Recommend Projects