'Name error' while calling function

Hi, simply move the definition "def plotAddData():" before your engines. The function needs to be defined before calling it.

HTH
Klaus

Hi Klaus,

Thanks for a prompt reply. I moved the function definition block before the engine declaration block as per your suggestion and executed the code.

The problem however still persists. The error message appearing on the terminal is as follows:

NameError Traceback (most recent call last)
/usr/bin/yade in <module>()

NameError: name 'plotAddData' is not defined

Any other suggestion?

Subha

Hi,

I didn't run the script previously but when looking at it in more detail now I can see there are a couple of issues. Below is a running version.

HTH
K

from yade import qt
from yade import plot,pack,geom

#define material properties:

rho = 2500 #density
fric = 0.5 #friction coefficient
fric_angle = atan(fric) #friction angle in radian
stiff_n = 1e8
stiff_s = 1e8
c_n = 0.1 #normal viscous damping
c_s = 0.1 #shear viscous damping
l_d = 0.2 #local damping

#define a material:

Mat = O.materials.append(ViscElMat(kn=stiff_n, ks=stiff_s, \
cn=c_n, cs=c_s, density=rho, frictionAngle=fric_angle))

#create a sphere:

b1 = utils.sphere(center=(0,0.0,1),radius=0.05,fixed=False,material=Mat)
O.bodies.append(b1)

#O.bodies.append(utils.sphere(center=(0,0,1), radius=0.05, material=Mat))

#create a plane facet:

O.bodies.append(geom.facetBox(center=(0,0,0), extents=(0.5,0.5,0), fixed=True, material=Mat))

# define data collector
def plotAddData():
 pos = b1.state.pos
 nIntrs = len( b1.intrs())
 plot.addData(i=O.iter,t=O.time,x=pos[0], y=pos[1], z=pos[2], Ek=utils.kineticEnergy(), nIntrs=nIntrs)

#define engines:

O.engines=[
 ForceResetter(),
 InsertionSortCollider([Bo1_Sphere_Aabb(),Bo1_Facet_Aabb()]), #Axis-aligned bounding box
 InteractionLoop(
  [Ig2_Sphere_Sphere_ScGeom6D(),Ig2_Facet_Sphere_ScGeom()], #Interaction geometry
  [Ip2_ViscElMat_ViscElMat_ViscElPhys()], #Interaction physics
  [Law2_ScGeom_ViscElPhys_Basic()], #Contact law
 ),
 NewtonIntegrator(gravity=(0,0,-9.81),damping=l_d),

# qt.SnapshotEngine(fileBase='/home/subhasish/YADE/tutorials/image/',label='snapper',iterPeriod=100),
 PyRunner(iterPeriod=5,command='plotAddData()')
]

#show geometry:

v=qt.View(); v.axes=True
v.viewDir=Vector3(0,1,-.05)
v.eyePosition=Vector3(0,-5,.5)

O.dt = 1e-4

#O.dt=.5*utils.PWaveTimeStep()

# wait=True will cause not returning to python until simulation will have stopped
O.run(50000,True)

# run this outside pyRunner for performance
plot.saveDataTxt('file2.txt.bz2')

Hi Klaus,

Thanks again to have a look at this. Sadly, the problem still persists. The outcome is:

ameError Traceback (most recent call last)
/usr/bin/yade in <module>()

NameError: name 'plotAddData' is not defined

and no data were written in the "txt" file.

I copy-pasted the modified code as it is and executed using both "yade" (ver. 1.07) and "yadedaily" (ver. 1.20) through the command execfile('filename.py').

Since you checked the code running at your end, I'm thinking about any missing "yade" component during installation. What would be your thought on this?

Many thanks.

Subha

Thanks Klaus!

It worked when executed with yade "filename.py".

Just wondering what is the difference when the script is run using execfile("filename.py") and yade "filename.py" because the former one works fine when no function is added.

Just another comment, I wanted to save both snapshots and data at the same time using PyRunner and for that purpose, I added the following lines to the script:

O.engines+=[PyRunner(iterPeriod=100,command='saveImg()'),
 PyRunner(iterPeriod=5,command='plotAddData()')]

and defined the image saving function as:

def saveImg():
 qt.SnapshotEngine(fileBase='/home/subhasish/YADE/tutorials/image/img-')

The geometry is shown in runtime as before:

#show geometry:

v=qt.View(); v.axes=True
v.viewDir=Vector3(0,1,-.05)
v.eyePosition=Vector3(0,-5,.5)

The script, however, does not save any runtime image.

Any suggestion will be appreciated.

Regards,

Subha

Thanks Klaus Thoeni, that solved my question.

Hi,

by using execfile not all yade modules are imported correctly.

Regarding your SnapshotEngine, just use it the way you had it before and make sure the directory you want to write into exists.

O.engines=[
 ForceResetter(),
 InsertionSortCollider([Bo1_Sphere_Aabb(),Bo1_Facet_Aabb()]), #Axis-aligned bounding box
 InteractionLoop(
  [Ig2_Sphere_Sphere_ScGeom6D(),Ig2_Facet_Sphere_ScGeom()], #Interaction geometry
  [Ip2_ViscElMat_ViscElMat_ViscElPhys()], #Interaction physics
  [Law2_ScGeom_ViscElPhys_Basic()], #Contact law
 ),
 NewtonIntegrator(gravity=(0,0,-9.81),damping=l_d),
 qt.SnapshotEngine(fileBase='your-sub-directory/img-',label='snapper',iterPeriod=100), # set directory
 PyRunner(iterPeriod=5,command='plotAddData()')
]

In addition, you have to set wait=False when you run it:

O.run(50000,False)

or just use the controller/python interface to run the simulation.

K

Regarding #6, I think your problem arose from variable scope (global or local) -- rather than modules issues.

Python variables / functions may have different scope (leading to possibility or impossibility to access them) depending how you launch your python script.

I bet that your python "plotAddData" function had a local scope when using execfile(...), as opposed to a global one when using yade yourScript.py

There has been previous (more precise) discussion on such topic in the mailing list before, but I can not tell you where/when...

Hi Klaus,

I noticed that if O.run(N,False) is used, the script does not write the data in the text file, but saves the images.

SM

Hi Jerome,

Thanks for your comment. How do I define the scope of a function local or global in the script?

Thanks,

SM

add #10: you have two options:
1. execute the line plot.saveDataTxt('file2.txt.bz2') in the terminal after you have finished the simulation
2. put the line back where you had it at the beginning, it was only a suggestion to move it outside the PyRunner for better performance

K

Thanks Klaus,

For some reason, option#1 resulted in a segmentation fault (core dumped) however option#2 worked but obviously at the expense of increased simulation time as you previously indicated.

SM

As for # 11, I finally found the previous discussion I had in mind. See https://answers.launchpad.net/yade/+question/239546, which should tell you how to directly define variables / functions scope.

Thanks Jerome. That post was very helpful indeed.