I am trying to convert all the coordinates of atoms within the simulation box of gromacs trajectory into a fractional coordinate by multiplying it with the inverse of a matrix.
I am using Fortran language and a libgmxfort library to read in the gromacs trajectory but the main issue that I am facing during this was storing coordinates (x,y,z) into separate variables as libgmxfort lib allows to store all the variables into a 3*n array using " a = trj%x(i1,j1) ", and separating x, y, z into individual variable from "a" is a task to further proceed.
Also, I tried " trj%frameArray(i1)%xyz(1,j1) ", which is not providing the correct values of coordinates for few of the atoms.
Below, I am attaching the code.
I will be grateful, if anyone can help regarding the above mentioned issue.
Anupama Sharma I'm not sure what the conversion is. The attached code works for the coordinate part.
Sun Tiedong ,yes it is working, but the values that is obtained using "trj%frameArray(i1)%xyz(1,j1)" is not matching with some of the values obtained using gmx dump command. But using "trj%x(i1,j1)" to get the coordinate does the work, but I need to seperate the "X","Y", and "Z" coordinate into individual variables from array " a " (where a=trj%x(i1,j1)) for further calculations.
Anupama Sharma I don't see why you need to separate the coordinate to individual component. You get a vector from "trj%x(i1,j1)". Then just multiply it by your matrix with matmul(). The problem is "rcell" is supposed to be a matrix. Why represent it with a vector?
https://gcc.gnu.org/onlinedocs/gfortran/MATMUL.html
Can you give the "gmx dump" command you used?
Anupama Sharma The code in my last screenshot works correctly on my machine. If you're looking at the data print out on your screen, remember those are after the conversion/transformation.
- Badges
- Science topic
- Similar topics
- Physical Sciences
- Atomic, Molecular and Optical Physics
- Atoms