Determine Spatially Resolved Magnetization

[Ponsudana]

Hello Ubermag Team,

How can I determine the spatially resolved magnetization at each cell for the magnetization components at all timesteps?
When I time drive a system for 20 ns and saved the data at a timestep of 5 ps (n=4000), the ODT file containing magnetization values mx, my and mz for all timesteps is obtained.
Now, I wish to determine the spatially resolved magnetization at each cell and save it as an ODT file.
So, I have to read the magnetization from all ODT files (4000 files) and determine the spatially resolved magnetization at each cell.
If region is p1=(-80e-9, -80e-9, 0), p2=(80e-9, 80e-9, 1e-9) and cell is (1e-9, 1e-9, 1e-9), the number of cells are 160 * 160
Now how to write the code to determine spatially resolved magnetization (mxs, mys and mzs) and save it as an ODT file.

Thank you,
Regards,
Ponsudana

[marijanbeg]

Hi @Ponsudana, thank you for your question. ODT files contain only spatially averaged magnetisation at all time steps. To obtain magnetisation values at all individual time steps, OMF files should be read - micromagneticdata is the package that can be used for that. Unfortunately, it is not documented fully yet, but the basic usage is:

import micromagneticdata as md

data = md.Data(system_name) # system_name is the name you gave to your system
drive = data[N] # N is the drive number in which you used TimeDriver
# Now, the magnetisation field at time_step at point can be accessed as:
drive[time_step](points)

@swapneelap has been working on some very nice functionalities for using xarray - maybe he can provide further information,.

[swapneelap]

Hello @Ponsudana and @marijanbeg,

With the next ubermag release, it will be possible to obtain an xarray.DataArray from discretisedfield.Field. Taking the same example as above:

import micromagneticdata as md

data = md.Data(system_name)  # system_name is the name you gave to your system
drive = data[N]  # N is the drive number in which you used TimeDriver
# Now, the magnetisation field at time_step at point can be accessed as:
mag_c = drive[time_step].to_xarray().sel(x=<c_x>, y=<c_y>, z=<c_z>, method=nearest)

The values of <c_x>, <c_y>, and <c_z> should be the coordinates (in meter) of the cell. mag_c.values should give an numpy.ndarray containing the magnetization components. One can also extract the magnetization components from cell indices by using to_xarray().isel instead of to_xarray().sel. Taking the example further:

import xarray as xr

collect_xr = []
for step in range(drive.n):
    collect_xr.append(drive[step].to_xarray())

drive_xr = xr.concat(collect_xr, dim='t')  # Collecting all DataArrays in one

mag_c_t = drive_xr.sel(x=<c_x>, y=<c_y>, z=<c_z>, method=nearest)

mag_c_t is an xarray.DataArray which contains the magnetization components of the cell at all the time steps.

We are also working towards adding to_xarray method to micromagneticdata.Drive which will yield the concatenated DataArray directly. This will also be included in the next release.