Data Access

After creating the simulation object all simulation data can be accessed via this object. Here the different ways to access data are explored.

If not explicitly specified, all examples assume an simulation object sim is available

import plutoplot as pp

sim = pp.Simulation("path/to/simulation")

The data for the individual timesteps is encapsulated in PlutoData-object, which the Simulation-object will create on demand. The data arrays for the variables can be accessed from there.

Accessing Steps

Direct access

To get the PlutoData-object for a specific step can be accessed using the []-operator or get() function like this:

step_data = sim[10]
step_data = sim.get(10, keep=False)
# or direct acces to variables:
sim[10].rho

get() also has an additional parameter keep, which chooses if the resulting object should be kept im memory after use.

Simulation.get() docs

plutoplot.simulation.Simulation.get(self, key, keep=True)

Get PLUTO output

Access individual data frames as PlutoData object. If keep=True the object is cached.

Parameters:

Name	Type	Description	Default
key	int	Output number	required
keep		obj:bool, optional): Keep output step in memory	True

Returns:

Type	Description
	obj:Simulation.DataObject (by default: :obj:PlutoData)

Iteration and reduce

It is also possible to iterate over all (or some steps for processing)

for step_data in sim:
    # process data here

# or for more control over the range
for step_data in sim.iter(start, stop, step):
    # process data

Simulation.iter() reference

plutoplot.simulation.Simulation.iter(self, *range_, *, keep=False)

Iterate over simulation

Range argument definition the same as range().

Attributes:

Name	Type	Description
simulation	plutoplot.simulation.Simulation	Simulation to iterate
keep	bool	Whether to keep the PlutoData objects in memory
start	int	Iteration start
stop	int	Iteration stop (exclusive)
step	int	Iteration step

Returns:

Type	Description
SimulationIterator	SimulationIterator

If each step should be reduced to into an array of values the reduce and reduce_parallel helpers can be used. For example, if the mean density of each step is needed:

mean_rho = sim.reduce(lambda step: step.rho.mean())
# result: mean_rho.shape == len(sim)

More complex reductions with n-dimensional outputs are also supported. The shape of the output can be either given as an argument, or will be deduced from the first calculation:

column_density = sim.reduce(lambda step: step.rho.sum(axis=2))
# Result: column_density.shape == (n1, n2)

reduce_parallel is equivalent, but will use multiprocessing to parallelise this operation. Be aware that because of the overhead of multiprocessing this is not necessarily faster than a serial reduce(), especially with an optimized numpy-installation.

reduce() reference

plutoplot.simulation.Simulation.reduce(self, func, range=(), dtype=None)

Reduce all simulation steps with function.

Shape and dtype are implied from return value of func().

Parameters:

Name	Type	Description	Default
func	function	function which takes a PlutoData object and returns scalar or numpy array.	required
dtype	numpy.dtype	forced data type for result array (if None func() implies dtype)	None
range	tuple	range tuple for iterator.	()

Returns:

Type	Description
numpy.ndarray	reduced data array

reduce_parallel() reference

plutoplot.simulation.Simulation.reduce_parallel(self, func, range=(), processes=None, dtype=None)

Reduce all simulation steps with function in parallel

Shape and dtype are implied from return value of func().

Parameters:

Name	Type	Description	Default
func	function	function which takes a PlutoData object and returns scalar or numpy array.	required
dtype	numpy.dtype	forced data type for result array (if None func() implies dtype)	None
range	tuple	range tuple for iterator.	()

Returns:

Type	Description
numpy.ndarray	reduced data array

Accessing data

The available data variables of a simulation can be accessed from simulation.vars or plutodata.vars: