search

materialsinnovation / pymks

Materials Knowledge System in Python
https://pymks.readthedocs.io/
Other
115 stars 76 forks source link

Reconstructing/recovering the microstructure back from 2PC stats. #567

Open Nimal136 opened 3 years ago

Nimal136 commented 3 years ago

Hi Team, I have a relative loading map of 2PCs after PCA and thought of reconstructing it back to the raw microstructure to look at how the reduced-order representation of the set of microstructures looks like. I remember reading this paper (doi.org/10.1016/j.actamat.2007.10.044) on recovering the microstructure from the computed 2PC stats using a phase recovery algorithm. Has this been implemented in the package or is there any plan to do it in future versions? Such functionality would be helpful to visualise what are the interesting features in the ensemble of microstructures. Cheers

wd15 commented 3 years ago

This is not implemented in PyMKS currently. No plans as yet either. If you do have a go at implementing it then I would certainly be interested in integrating that into PyMKS.

Edit: if you have a prototype implementation then we can work that into PyMKS

owaisahmad18 commented 1 year ago

data_correlation = TwoPointCorrelation( periodic_boundary=True, cutoff=25, correlations=[(0,0),(0,1)] ).transform(data)

how can i use the output from TwoPointCorrelation command in pymks package and feed it to MCRpy or any other package to reconstruct microstructure. please help.

wd15 commented 1 year ago

@owaisahmad18 the output from PyMKS is just a Numpy array which you can save in a variety of formats. I suspect that MCRpy would also be using Numpy arrays so you could use the two tools in a single Python session.