Lexachoc
commented
6 months ago @ChemikLopaton Hi, Unfortunately, this repo only contains the experimental data and the predicted properties are not included.
That's something I thought was a pity. However, I think this is also an opportunity to encourage researchers/scientists to develop advanced predictive models. Priven2000 is over 20 years old and I think it is time to improve those models in the SciGlass.
Lots of articles about ML models and Python tools like https://github.com/drcassar/glasspy (I haven't tried it yet).
Therefore, I create a web platform of SciGlass (see https://sciglass.uni-jena.de/) to integrate these useful ML models for researchers/scientists. Looking for collaboration for models!
Hello, I have a question that might be a bit naive, but please, in this database, are the computational capabilities for predicting glass properties still retained? I have tried searching and most resources talk about the old sciglass software which seems to be discontinued (but still referenced in publications) as I got no response multiple times form the official sciglass website and have not found any way of obtaining the software. Is there any documentation to the database provided in this repository? To make the question more concrete, I am talking about this part of the feature list: Property Calculations. Over 100 computational methods to compute the properties in 15 groups (e.g., viscosity, density, mechanical, optical), many of them in broad temperature ranges. Prediction of properties of oxide, halide, and chalcogenide glasses is possible in wide concentration ranges. as I was hoping to use sciglass to calculate viscosity curves and melt surface tension for relatively simple glass systems. I just have hard time navigating the database or finding any relevant documentation regarding the db structure or the use/specification of the computational methods.