Quick start
Visit the Installation page to install phyddle.
Download a copy of the phyddle repository to your workstation either as .zip file or by cloning the git repository.
To run a phyddle analysis enter your new directory, e.g.:
# enter phyddle directory
cd ~/projects/phyddle
Several example phyddle analyses are stored in the ./workspace
subdirectory.
Each project directory contains a Configuration file, a
Simulation script, and will contain directories for output
resulting from an analysis. For example, you can run an analysis for the
bisse_r
project using the commands:
# enter project directory
cd ./workspace/bisse_r
# run phyddle with project config
phyddle -c config.py
This will run an analysis using the default settings for all Pipeline
steps. As the first step, Simulate will execute the command string
stored under sim_command
in the config file. For ./workspace/bisse_r/config.py
this command is
# phyddle runs this R script to simulate
Rscript ./sim_bisse.R
Note
You must be able to run the simulation script from command line without
administrative privileges. This means must install the required software
for each project. The above script requires R and the packages ape
and castor
to run.
Eventually, you will want to write your own config file. There are two easy
ways to create your own config file. One option is to copy and modify an
existing script. Another option is to create a new blank config with
phyddle --make_cfg
and then modify the new file
# create a default config file
phyddle --make_cfg
# rename the config file
mv config_default.py new_config.py
# update the config file
edit new_config.py
Now, let’s assume new_config.py
specifies a phyddle analysis with 1000 simulated
training examples, using R for simulation. Run phyddle against the new config
with the following command:
# run phyddle with new config file
phyddle -c ./new_config.py
Provide phyddle with command-line options to customize how each pipeline step is executed. Visit Pipeline and Workspace to learn more about managing phyddle analyses.
In practice, you’ll want to generate a larger training dataset with anywhere from 10k to 1M examples, depending on the model. To add new examples to your training set, for example:
# enter project directory
cd workspace/my_project
# [S]imulate new training examples, stored in
# ./simulate subdirectory
phyddle -s S -c config.py --sim_more 14000
# [F]ormat all raw_data examples as tensors,
# stored in ./format subdirectory
phyddle -s F -c config.py
# [T]rain network with tensor_data, but override batch size,
# stored in ./train subdirectory
phyddle -s T -c config.py --trn_batch_size 256
# [E]stimate parameters for biological dataset, with results
# stored in ./estimate subdirectory; and then [P]lot
# figures, storing them in ./plot subdirectory
phyddle -s EP -c config.py
Visit Configuration to learn more about currently supported phyddle settings. View supported command-line options with:
# see help for phyddle
phyddle --help