SÃ¸ren. Error in Anybody v3.0 with files "txt" ...

system · January 3, 2008, 9:16pm

Hi again SÃ¸ren,
Excuse me, but we have got to change the 53 “txt” files, at the
moment, manually with the new configuration (only 1 line in each
file).
Then when we try to run AnyBody ver.3.0, but we get the following
error:

ERROR(SCR.EXP10) : D:__AnyBody_LEMUS_Rep6.1
\BRep\Aalborg\Scaling\ScalingLengthMass.any(105) : ‘ScaleMat’ :
Expression evaluation failed at moment ‘Const’ :
D:__AnyBody_LEMUS_Rep6.1\BRep\Aalborg\Scaling\ScalingLengthMass.any
(103) : ‘ls’ : argument will not be ready for evaluation until
moment ‘ConfigVar’

We will send you the folders which include the “any” and “txt” files.

The folder “Aalborg” only includes the directories modified from
standard repository 6.1.

The file name will be: AnyBody.3.0_files_txt_JavierMarin_2008Jan03.rar

We don’t know what is wrongâ€¦

Please, could you help us?
Thanks in advance,

Javier Marin.
Ergonomics Group of University Zaragoza (SPAIN)

system · January 4, 2008, 1:40pm

Hi Javier

The error message you get please see below, means that the ScaleMat makes
use of a variable “ls” which is not ready to be used. The variable “ls” is
calculated as a function of the “ThighLength” which has been read in by the
AnyInputFile statement. The problem with this is that reading the variable
in this way makes the variable “ThighLength” a ConfigVar which is not ready
to be used yet when the ScaleMat is being calculated as a Const. AnyScript
operates with different categories of variables and these are calculated at
different moments when the model is being loaded, this makes it impossible
to assign a Const variable like the ScaleMat using a variable like
ThighLength because when the ScaleMat is being calculated the ThighLength
has not been found yet in the file, because it has a later evaluation
moment, since it is a ConfigVar. Please see the reference manual page 12
section 1.6 for details, especially page 14 list the some of the different
types of variables.

Error Message:

ERROR(SCR.EXP10) :
C:\stc\CustomerZ\Repository.6.1\BRep\Aalborg\Scaling\ScalingLengthMassFat.an
y(94) : ‘ScaleMat’ : Expression evaluation failed at moment ‘Const’ :

C:\stc\CustomerZ\Repository.6.1\BRep\Aalborg\Scaling\ScalingLengthMassFat.an
y(92) : ‘ls’ : argument will not be ready for evaluation until moment
‘ConfigVar’

This effectively makes it impossible to make use of the AnyInputFile object
for reading in the files, where the variables are used for scaling the
model. I am very sorry that i did not foresee this problem, when I suggested
that solution. It did not cross my mind that it could be a problem, but this
way of reading the variables is not the normal way we do this. Now I realize
that the only way to read in the anthropometric variables which are going to
be used for scaling is to have them written as AnyScript files.

I have modified the file Main_Anthro.any and added an extra include file
“AnyMan.any”

This include files looks like this :

AnyVar BodyHeight =1.97;

AnyVar ThighLength = 0.559;

AnyVar ShankLength = 0.424;

AnyVar FootLength = 0.217;

AnyVar PelvisWidth =0.21;

AnyVar TrunkHeight = 0.712;

AnyVar HeadHeight = 0.189;

AnyVar UpperArmLength = 0.312;

AnyVar LowerArmLength = 0.289;

This is the normal way we read in anthropometric parameters.

In the Main file there was an identical problem with the tStart, tEnd and
nStep values, for this purpose I created the file “Times.any” which contains
these variables.

Finally I discovered that in almost all the files used for defining the
motion there was an extra empty line in the files, this caused a problem for
the AnyInputFile reader so it will only return the first number. After
having removed the extra lines in all the files the model loaded ok. You can
read more about the AnyInputFile in the reference manual on page 74. The
AnyInputFile object is originally designed for reading in several columns
where the first columns is time values and the subsequent columns are
positions or similar. It is not designed for the way we are using it here,
for reading in one vector, but this is the only way to read in the files you
have. I think this unintended use of the object causes the sensitive to the
extra line, it could be error prone. The best way to read in the data would
be to have both the time vector and the value vector in the same file, if
your file setup does allow this. So if it is possible to modify the
application which creates these files this would be the most secure
solution.

I will upload a new version of the model with the changes implemented to the
FileSection of the group, shortly.

Best regards

SÃ¸ren, AnyBody Support

From: anyscript@yahoogroups.com [mailto:anyscript@yahoogroups.com] On Behalf
Of jjmarinz
Sent: 03 January 2008 21:16
To: anyscript@yahoogroups.com
Subject: [AnyScript] SÃ¸ren. Error in Anybody v3.0 with files “txt” …