[velleri3mainz]

Hello,
I hope you can help me.

I want to read the xyz values for the target position of a movement from a program.
The program was generated from a CAM file and is functional without errors in RoboDK.

Example:

#Example move command
instruction_dict = program.setParam(5)

print(instruction_dict)
#{'Joints': '109.507535,-13.721288,122.115482,-115.355384,92.927156,-78.386586', 'MoveType': 1, 'Name': 'MoveJ 1', 'Pose': 'transl(-298.576,-198.535,136)*rotx(-180)*rotz(-4.981)', 'TargetPtr': '0', 'Type': 0}

print(instruction_dict['Pose'])
#transl(-298.576,-198.535,136)*rotx(-180)*rotz(-4.981)

print(str(Pose_2_TxyzRxyz(instruction_dict.Pose()))) #AttributeError: 'dict' object has no attribute 'Pose'

#what is missing to get x y z?

print(x)# -298.576
print(y)#-198.535
print(x)#136


best regards
Jonas Veller

[Albert]

It is better if you use the Instruction function:
https://robodk.com/doc/en/PythonAPI/robo...nstruction

So you can do:

Code:

ins_name, ins_type, move_type, isjointtarget, pose, joints = prog.Instruction(0)
if ins_type == INS_TYPE_MOVE:
    print(pose)

You could also use the eval function on the pose string. However, if you use the global robomath context you should convert degrees to radians. You can create a custom context for that specific eval function. Example:

Code:

from robodk import robomath
import math
context = {}
context["transl"] = robomath.transl
context["rotx"] = lambda rx: robomath.rotx(rx * math.pi / 180.0)
context["roty"] = lambda ry: robomath.roty(ry * math.pi / 180.0)
context["rotz"] = lambda rz: robomath.rotz(rz * math.pi / 180.0)
# pose = eval("transl(100,0,0)*rotx(180)*roty(90)", context)
pose = eval(instruction["Pose"], context)
print(pose)

[velleri3mainz]

Perfect that has solved my problem:

ins_name, ins_type, move_type, isjointtarget, pose, joints = prog.Instruction(0)

Thank you

[velleri3mainz]

Hello again,
the described way about

ins_name, ins_type, move_type, isjointtarget, pose, joints = prog.Instruction(0)

works for MoveJ instruction.
Now I have come across a MoveC instruction, where I unfortunately get no further.
Unlike MoveJ Instrucktion, MoveC does give me only poses in the as Strings.


instruction_dict = program.setParam(ins_id)
print(instruction_dict)
#{'Joints': '103.670178,-13.553023,108.458609,-121.166389,87.567876,-90.806867', 'Joints2': '103.675661,-13.502296,108.406457,-121.160885,87.566737,-90.807845', 'MoveType': 3, 'Name': 'MoveC 7', 'Pose': 'transl(505,803.679,19.343)*rotx(180)*rotz(-4.981)', 'Pose2': 'transl(505,802.972,19.050)*rotx(-180)*rotz(-4.981)', 'Type': 1}

print(program.Instruction(ins_id))
#('MoveC 7', 1, None, None, None, None)

Do i miss something or is using the specific eval function from Albert the only way?

Best regards
Jonas

[Albert]

The other method I mentioned with the setParam function offers more flexibility such as dealing with circular movements and many other features.

The code I shared in my first response using eval should work. Did you try it?

You can create a function to facilitate reusing this code. Example:

Code:

def String2Pose(posestring):
    from robodk import robomath
    import math
    context = {}
    context["transl"] = robomath.transl
    context["rotx"] = lambda rx: robomath.rotx(rx * math.pi / 180.0)
    context["roty"] = lambda ry: robomath.roty(ry * math.pi / 180.0)
    context["rotz"] = lambda rz: robomath.rotz(rz * math.pi / 180.0)
    # pose = eval("transl(100,0,0)*rotx(180)*roty(90)", context)
    pose = eval(posestring, context)
    return pose

pose = String2Pose(instruction["Pose"])
pose2 = String2Pose(instruction["Pose2"])

[velleri3mainz]

Thanks Albert the Code you shared works much better.