I’ve been sucessful in using the application for human gait simulation on various datasets with good results and no problems for adults, but I have new datsets collected from 2 children that I need to analyze and it’s proving difficult.
I am using 15 markers, similar to the original application, only with a added SACR marker and removal of LPSI and RPSI
I have tested the data from the markers and it looks very good and has no visible errors regarding the x,y,z positioning.
I have scaled down the human model to appropiate size and positioned it closely to the markers, the positioning is not perfect, but close(still in the test phase and waiting for exact measures).
I have not made any changes to the file that writes output for the gaitapplication2 (DataForConfigFile.any)
The problem is as follows:
After using the marker position generation study (the first one that returns RTOE.txt, etc), I run the second study for making an output for gaitapplication2 program, all seems well, but when I execute, I get this for all time steps “Objective function value: -1.#IND0000000000, at time: …”, it says that the operation was a sucess, but obviously this is not the case.
I am running the application with the following imput(the standard one):
First of all please consider using the new Ver. 4.1 for this kind of analysis. In this version the kinematic optimization is a built in feature so you do not need the gaitapplication2 anymore, this makes the process much easier…
I am not sure what exactly goes wrong in this case but here are some ideas:
As i understood it you have slightly changed the marker configuration around the hip but you write that the DataForConfigFile is unchanged, this sounds strange if you change the marker configuration i think this should also be change in this file accordingly.
I apologize for not being clear on what exactly I did to the code. I had also changed everything so the new marker configuration fits on the body. I made new tests starting from scratch and ended up with the same result, this time, I removed the two markers LPSI and RPSI and left the rest of the code to resemble the original as much as possible. Same eroor.
Other testing I made was to use a ModelGait.txt from the original application GaitTD along with my marker data(RTHI.txt, LASI.txt, etc) and time modifications required. This gave me the same error. What is strange is that if I use the original marker coords and the exported ModelGait.txt using the smaller bone configuration, I get normal error, with a numerical value rather than -1.#IND00… . However, the markers I have collected move in the right way on the screen and are in their intended anatomical position (i.e LTOE above the left toe).
As for the suggestion to try and use Ver 4.1, I have tried using the LowerExtremity model but I have no idea of how to get the data in the program since it is not in .c3d format and I actually made a convertor that gets x,y,z data from .txt files in the right format for GaitUniMiami to read (example: trial02.Header.Any is generated by my program and needs just to be copied). I have tried to trick the application into reading my files by adding #include statements for them and running, but the obvious already declared folders and variable references exist in the memory and I get a parsing error. Another attempt was made with an empty file and #include for all my files, but it failed because the header was not C3d and AnyBody recognized this and stopped the run.
Anyway, I hope I have given enough details to make the problem more clear.
As a sidenote, the routine I’m using for generation of files for the child gait has been used succesfully for adults with good results and no problems, and I can always verify the output visually in AnyBody or matlab which it was built in, so I think the problem is not that, but something else that escapes me right now.
I do not have many ideas on what goes wrong with the gaitapplication2 sorry, it is not an error i recall to have seen often. My only idea is that it could be related to the direction of walking, is this different from what it has been in the other datasets? if so please change the coordinate system so that it resembles the one in GaitUniMiami, i think the gaitapplicaion2 is limited in this way.
Concerning the c3d data:
I am sure there is a way of applying the data to v4.1 without having the c3d file, some users are doing this. It can be done in two ways:
1 use the new model GaitLowerExtremity model as basis and recreate the structure of the c3d file by reading in the data from the files in a folder structure resembling the one the c3d object would create. The the model will be able to find all the data as if the AnyInputC3d object has been used.
2 use the same model as basis and read in the all the data into data structures and rearrange the model to cope with a different data structure than the one which resemble c3d.
I discovered the problem, and it was related to a portion of code reffering to markers that did not exist, it was in trialXX.PointsMarker.any(it was confusing because all export files seemed to be in order). Now everything is working fine.
In regards to the suggested methods, I’m going to try the first one because it’s the most logical, and probably write an application that converts the data that I already have in xyz to a format readable directly by AnyBody.