Wavefront Corrector

In pyRTC, one of the core components is the wavefront corrector object. It typically finishes the AO chain by continously waiting for new corrections and applying them. This object is a consumer of the wfc shared memory object and a producer of the wfc2D shared memory objects which contain the current correction vector and the 2D representation of the curent correction respectively. This class required you to properly lay out the 2D actuator layout as well as define the correction basis.

Soft-RTC Example

The following is an example of how to initialize a WavefrontCorrector component in pyRTC.

Here we are in the soft-RTC mode of pyRTC, which holds all components in the same python process. See below for how to launch a hard-RTC equivalent.

"""
First we import the relevant wavefront corrector class. Typically, this will be a
specific hardware class which has been defined to work with the SDK of your corrector.

As an example (see hardware/ALPAODM.py):

from pyRTC.hardware import ALPAODM

Here, I will just initialize the Wavefront Sensor Superclass as an example
"""

#%% Run in interactive python or jupyter notebook to keep process alive
from pyRTC.WavefrontCorrector import WavefrontCorrector
import matplotlib.pyplot as plt
from pyRTC.utils import read_yaml_file

confWFC = {
"name": "example",
"numActuators": 97,
"numModes": 50,
"m2cFile": "", #Here you put the path to your basis ([nAct,nMode]) ./EXAMPLE/calib/wfcShape.npy"
"saveFile": "", #Here you put where the WFC will save its corrections ./EXAMPLE/calib/wfcShape.npy"
"affinity": 2,
"functions": ["sendToHardware"]
}

"""
Alternatively, read the config from a file

conf = read_yaml_file("./EXAMPLE/config.yaml")["wfs"]
"""

#Initialize the WFS object
wfc = WavefrontCorrector(confWFC)
#Start the functions regiserted to the loop (i.e, expose)
wfc.start()

wfc.flatten()

Hard-RTC Example

The following is an example of how to initialize a WavefrontCorrector component in pyRTC.

Here we are in the hard-RTC mode of pyRTC, which holds all components in the separate python processes. This circumvents the python Global Interpreter Lock.

See above for how to launch a soft-RTC equivalent.

from pyRTC.Pipeline import hardwareLauncher

"""
For the Hard-RTC, you will need to set-up a config before hand and store it in a yaml file.

It should look something like:

wfc:
  name: "ALPAO"
  serial: "BAX118"
  numActuators: 97
  numModes: 94
  flatFile: "./examples/sharp_lab/calib/wfcShape.npy"
  saveFile: "./examples/sharp_lab/calib/wfcShape.npy"
  m2cFile: "./examples/sharp_lab/calib/m2c_kl.npy"
  affinity: 5
  commandCap: 0.8
  hardwareDelay: 0.001 #seconds
  frameDelay: 0
  functions:
  - sendToHardware
"""
config = 'path/to/config.yaml'
port = 3000

#Initialize the hardware launcher for your WFS child hardware class
wfc = hardwareLauncher('path/to/pyRTC/hardware/ALPAODM.py',config,port)

"""
Launch the process.

This will run the hardware file, which should establish a connection with the current process.
This is accomplished with the Listener class (see hardware folder for examples).

The functions registered in the config to the real-time loop will automatically be started.
"""
wfc.launch()

"""
Once the connection has been made successfully, you can run any function in the hardware class
using the run function. You can also get and set properties of the hardware using getProperty()
and setProperty() respectively.
"""
wfc.run("flatten")

wfc.setProperty("commandCap", 0.6)

print(wfc.getProperty("commandCap"))

Parameters

class pyRTC.WavefrontCorrector.WavefrontCorrector(conf)

Bases: pyRTCComponent

Base class for deformable mirrors and other wavefront-correction devices.

WavefrontCorrector is responsible for the control-plane machinery around command generation: SHM output, flat shapes, mode-to-command transforms, floating actuator handling, and delayed command buffers. Subclasses are left to implement the device-specific transport in sendToHardware.

Config

namestr

Name of the wavefront corrector.

numActuatorsint

Number of actuators. Required.

numModesint

Number of modes. Required.

affinitystr

Affinity setting.

m2cFilestr

Path to the mode-to-command file.

floatingInfluenceRadiusint, optional

Radius for floating influence. Default is 1.

frameDelayint, optional

Frame delay. Default is 0.

saveFilestr, optional

File to save the shape. Default is “wfcShape.npy”.

Attributes

namestr

Name of the wavefront corrector.

numActuatorsint

Number of actuators.

numModesint

Number of modes.

affinitystr

Affinity setting.

m2cFilestr

Path to the mode-to-command file.

correctionVectorImageSHM

Correction vector.

correctionVector2DImageSHM or None

2D correction vector for display.

flatnumpy.ndarray

Initial flat shape.

flatModalnumpy.ndarray

Flat shape in modal basis.

currentShapenumpy.ndarray

Current shape.

actuatorStatusnumpy.ndarray

Status of each actuator.

index_mapnumpy.ndarray or None

Index map for actuators.

floatingInfluenceRadiusint

Radius for floating influence.

floatMatrixnumpy.ndarray

Floating actuator matrix.

frameDelayint

Frame delay.

saveFilestr

File to save the shape.

layoutnumpy.ndarray or None

Layout of the actuators.

M2Cnumpy.ndarray

Mode-to-command matrix.

f_M2Cnumpy.ndarray

Floating mode-to-command matrix.

C2Mnumpy.ndarray

Command-to-mode matrix.

currentCorrectionnumpy.ndarray

Current correction vector.

shapeBuffernumpy.ndarray

Buffer for shapes with frame delay.

correctionVector2D_templatenumpy.ndarray

Template for the 2D correction vector.

deactivateActuators(actuators)

Deactivate specified actuators. Actuators are assumed to be floating

Parameters

actuatorslist of int

List of actuator indices to deactivate.

flatten()

Flatten the wavefront corrector.

loadFlat(filename='')

Loads the Flat from a file.

Parameters

filenamestr, optional

Filename to load the dark frame from. If not specified, uses the dark file path from the configuration.

plot(addFlat=False)

Plot the current correction.

Parameters

removeFlatbool, optional

If True, removes the flat shape from the current correction before plotting. Default is False.

push(mode, amp)

Push a specific mode with a given amplitude.

Parameters

modeint

Mode index to push.

ampfloat

Amplitude to push the mode with.

reactivateActuators(actuators)

Reactivate specified actuators.

Parameters

actuatorslist of int

List of actuator indices to reactivate.

read(block=False)

Read the current correction vector.

Returns

numpy.ndarray

Current correction vector.

readM2C(filename='')

Read the mode-to-command matrix from a file.

Parameters

filenamestr, optional

File to read the mode-to-command matrix from. If not specified, uses the configured m2cFile.

saveShape(filename='')

Save the current shape to a file.

Parameters

filenamestr, optional

File to save the shape to. If not specified, uses the configured saveFile.

sendToHardware()

Send the current correction to the hardware. Nominally, this function is overwritten by the child hardware class and registered to the real-time loop from the config.

setDelay(delay)

Sets an artificial frame delay. Used for testing, nominally the delay should always be zero.

Parameters

delayint

Frame delay to set.

setFlat(flat)

Set the flat shape.

Parameters

flatnumpy.ndarray

Flat shape to set.

setLayout(layout)

Set the layout of the actuators.

Parameters

layoutnumpy.ndarray or None

Layout of the actuators. Is converted to boolean if not already.

setM2C(M2C)

Set the mode-to-command matrix. This is the basis for correction.

Parameters

M2Cnumpy.ndarray or None

Mode-to-command matrix to set. Axes are [numActuators, numModes]

start()

Start the registered real-time functions.

stop()

Stops the registered real-time functions.

write(correction)

Write a new correction.

Parameters

correctionnumpy.ndarray

Correction vector to write.