Skip to content

Simulate chemical reaction networks using ordinary differential equations.


Notifications You must be signed in to change notification settings


Folders and files

Last commit message
Last commit date

Latest commit



61 Commits

Repository files navigation


Simulate Chemical Recation Networks (CRNs) using Ordinary Differential Equations (ODEs).

GitHub tag (latest by date) GitHub release (latest by date including pre-releases) PyPI version PyPI - License Travis (.org) Codecov


Using the crnsimulator executable:

Create a test file with your CRN:

File: oscillator.crn

# Rock-Paper-Scissors Oscillator

A + B -> B + B [k = 0.2]
B + C -> C + C [k = 0.4]
C + A -> A + A [k = 0.7]

And pipe it into the crnsimulator:

~$ crnsimulator -o ozzy < oscillator.crn

This writes the ODE system to an executable python script:

Check the command line parameters of You have to set initial species concentrations, and choose an output-format, e.g.:

~$ python ./ --p0 A=0.1 B=1e-2 C=1e-3 --t8 10000 --pyplot ozzy.pdf

This example plots a simulation on a linear-time scale (0 - 10000) to the file ozzy.pdf .

Tips and Tricks:

You can pass the command line options for directly to crnsimulator. This will automatically simulate your ODE system. Use --force to overwrite an existing script.

~$ crnsimulator --p0 A=0.1 B=1e-2 C=1e-3 --t8 10000 -o ozzy --pyplot ozzy.pdf < oscillator.crn

You can specify the CRN in a single line:

~$ echo "A+B->2B [k=0.2]; B+C->2C [k=0.4]; C+A->2A" | crnsimulator --p0 A=0.1 B=1e-2 C=1e-3 --t8 10000 -o ozzy --pyplot ozzy.pdf

You can specify default initial concentrations of species:

~$ echo "A @i 0.1; B @i 1e-2; A+B->2B [k=0.2]; B+C->2C [k=0.4]; C+A->2A" | crnsimulator --p0 C=1e-3 --t8 10000 -o ozzy --pyplot ozzy.pdf

If you can set which species appear in the legend using --pyplot-lables. If you are writing a new executable (you may need --force), then you can also control the order:

~$ echo "A @i 0.1; B @i 1e-2; A+B->2B [k=0.2]; B+C->2C [k=0.4]; C+A->2A" | crnsimulator --p0 C=1e-3 --t8 10000 -o ozzy --pyplot ozzy.pdf --force --pyplot-labels C B

Using the crnsimulator library:

The easiest way to get started is by looking at the crnsimulator script itself. However, here is a small example using the above oscillating CRN.

>>> from crnsimulator import ReactionGraph
>>> crn  = [[['A', 'B'],['B','B'],0.2],
            [['B', 'C'],['C','C'],0.8],
            [['C', 'A'],['A','A'],0.9]]
>>> RG = ReactionGraph(crn)
>>> svars = ['B', 'C', 'A'] # let's enforce the order of species, because we can!
>>> filename, odename = RG.write_ODE_lib(filename='', sorted_vars = svars)
>>> print('Wrote ODE system file:', filename)
Wrote ODE system file:

Then go ahead and execute

~$ python ./ --p0 1=1e-6 2=2e-6 3=5e-6 --t8 1e8 --pyplot ozzy.pdf --atol 1e-10 --rtol 1e-10

... or load its functions by treating it as a python library:

# Import
>>> import numpy as np
>>> from scipy.integrate import odeint
>>> from crnsimulator import get_integrator
>>> odesys = get_integrator(filename, function = odename)
>>> odeplt = get_integrator(filename, function = 'ode_plotter')
# Simulate
>>> p0 = [1e-6, 2e-6, 5e-6] # order of svars
>>> time = np.linspace(0, 1e8, num = 10_000)
>>> ny = odeint(odesys, p0, time, (None,), atol = 1e-10, rtol = 1e-10).T
# Plot
>>> odeplt(`ozzy.pdf`, time, ny, svars)

... or include the prebuilt integrator in you own script (like the crnsimulator exectuable):

>>> from crnsimulator import get_integrator
>>> integrate = get_integrator(filename)
>>> integrate(args) # args = <argparse.ArgumentParser()>


~$ python install


v0.9 -- code cleanup

  • removed networkx dependency
  • moved plotting libraries and functions into a separate file to avoid automatic import.

v0.8 -- beta status

  • now using logging
  • python >= 3.7 only
  • improved header documentation
  • using entry_points for crnsimulator script
  • set defaultrate = 1 (new postprocessing strandard)
  • new commandline arguments: labels, labels-strict
  • support the constant concentration flag