Command line options¶
The full set of command line options for metawards is below;
usage: metawards [-h] [--version] [-i INPUT] [-l [LINE [LINE ...]]]
[-r REPEATS] [-s SEED] [-a [ADDITIONAL [ADDITIONAL ...]]]
[-o OUTPUT] [-d DISEASE] [-m MODEL] [--start-date START_DATE]
[--start-day START_DAY] [-p PARAMETERS] [-R REPOSITORY]
[-P POPULATION] [-n NSTEPS] [-u USER_VARIABLES]
[--iterator ITERATOR] [--extractor EXTRACTOR] [--UV UV]
[--nthreads NTHREADS] [--nprocs NPROCS] [--hostfile HOSTFILE]
[--cores-per-node CORES_PER_NODE] [--auto-bzip]
[--no-auto-bzip] [--force-overwrite-output]
[--max-nodes MAX_NODES] [--max-links MAX_LINKS] [--profile]
[--no-profile] [--mpi] [--scoop]
MetaWards epidemic modelling - see https://github.com/metawards/metawards for
more information
optional arguments:
-h, --help show this help message and exit
--version Print the version information about metawards
-i INPUT, --input INPUT
Input file for the simulation that specifies the
adjustable parameters to change for each run of the
model. You must supply some input to run a model!
-l [LINE [LINE ...]], --line [LINE [LINE ...]]
Line number (or line numbers) containing the values of
adjustable parameters to run for this run (or runs) of
the model If this isn't specified then model runs will
be performed for adjustable parameters given on all
lines from the input file. You can specify many
numbers, and ranges are also accepted, e.g. '-l 5 6-10
12,13,14' etc. Note that the line numbers are
0-indexed, e.g. the first line of the file is line 0.
Ranges are inclusive, so 3-5 is the same as 3 4 5
-r REPEATS, --repeats REPEATS
The number of repeat runs of the model to perform for
each value of the adjustable parameters. By default
only a single run will be performed for each set of
adjustable parameters
-s SEED, --seed SEED Random number seed for this run (default is to use a
random seed)
-a [ADDITIONAL [ADDITIONAL ...]], --additional [ADDITIONAL [ADDITIONAL ...]]
File (or files) containing additional seed of outbreak
for the model. These are used to seed additional
infections on specific days at different locations
during a model run
-o OUTPUT, --output OUTPUT
Path to the directory in which to place all output
files (default 'output'). This directory will be
subdivided if multiple adjustable parameter sets or
repeats are used.
-d DISEASE, --disease DISEASE
Name of the disease to model (default is 'ncov')
-m MODEL, --model MODEL
Name of the input model data set for the network
(default is '2011Data')
--start-date START_DATE
Date of the start of the model outbreak. This accepts
dates either is iso-format, or fuzzy dates such as
'monday', 'tomorrow' etc. This is used to work out
which days are weekends, or to make it easier to
specify time-based events.
--start-day START_DAY
The start day of the model outbreak. By default the
model outbreak starts on day zero (0), with each step
of the model representing an additional day. Use this
to start from a later day, which may be useful if you
want to more quickly reach time based events. Note
that the passed '--start-date' is always day 0, so day
10 has a date which is 10 days after start-date
-p PARAMETERS, --parameters PARAMETERS
Name of the input parameter set used to control the
simulation (default 'march29')
-R REPOSITORY, --repository REPOSITORY
Path to the MetaWardsData repository. If unspecified
this defaults to the value in the environment variable
METAWARDSDATA or, if that isn't specified, to
$HOME/GitHub/MetaWardsData
-P POPULATION, --population POPULATION
Initial population (default 57104043)
-n NSTEPS, --nsteps NSTEPS
Maximum number of steps (days) to run for the
simulation. Each step represents one day in the
outbreak (default is to run for a maximum of two years
- 730 days)
-u USER_VARIABLES, --user-variables USER_VARIABLES
Name of the file containing user-defined custom
variables. These provide extra information that can be
read by the custom integrators or custom extractors.
--iterator ITERATOR Name of the iterator to use to advance the outbreak at
each step (day). For a full explanation see the
tutorial at https://metawards.github.io
--extractor EXTRACTOR
Name of the extractor to use to extract information
during a model run. For a full explanation see the
tutorial at https://metawards.github.io
--UV UV Value for the UV parameter for the model (default is
0.0)
--nthreads NTHREADS Number of threads over which parallelise an individual
model run. The total number of cores used by metawards
will be nprocesses x nthreads
--nprocs NPROCS The number of processes over which to parallelise the
different model runs for different adjustable
parameter sets and repeats. By default this will
automatically work out the number of processes based
on the way metawards is launched. Use this option if
you want to specify the number of processes manually.
--hostfile HOSTFILE The hostfile containing the names of the compute nodes
over which to run a parallel job. If this is not set,
the program will attempt to automatically get this
information from the cluster queueing system. Use this
if the auto-detection fails
--cores-per-node CORES_PER_NODE
Set the number of processor cores available on each of
the compute nodes in the cluster that will be used to
run the models (if a cluster is used). If this is not
set then the program will attempt to get this
information from the cluster queueing system. Use this
option if the auto-detection fails.
--auto-bzip Automatically bz2 compress all output files as they
are written.
--no-auto-bzip Do not automatically bz2 compress all output files as
they are written.
--force-overwrite-output
Whether or not to force overwriting of any existing
output. Using this option will tell metawards that it
is safe to delete the contents of the output directory
specified in by '--output' if it already exists.
Dangerous as this can remove existing output files
--max-nodes MAX_NODES
Maximum number of nodes that can be read
--max-links MAX_LINKS
Maximum number of links that can be read
--profile Enable profiling of the code
--no-profile Disable profiling of the code
--mpi Force use of MPI to parallelise across runs
--scoop Force use of scoop to parallelise across runs