This project provides three functions for pyroxene's instances:
- calculator: convert oxide weight percentages data into molar ion concentrations;
- predictor: predict Fe3+/Fetot value with the given data set by trained ML models respectively;
- classifier: classify every entry into a specific kind of pyroxene according to the classification criteria of Commission on New Minerals and Mineral Names.(Morimoto, N. (1988). Nomenclature of pyroxenes. Mineralogy and Petrology, 39(1), 55–76)
See Manual_for_Mac or Manual_for_Windows for application details.The manual would tell:
- configuration
- data format
- package details
- operation steps
- model_training_code : this directory contains the codes to train the merchine learning model with designated hyperparameters and data sets used for train
- Fe3_calculation_program : this directory contains the codes packaged in 【pyroxene_processor】to provide three functions mentioned above and other related dependencies
- Can He (Sany)
Email : [email protected] - Weihua Huang
Email : [email protected]
- Huang Weihua, Lyv Yang, Du Minghao, He Can, Gao Shangde, Xu Renjun, Xia Qunke, ZhangZhou J* Estimation of ferric iron contents in clinopyroxene by machine learning models. American Mineralogist 107, 1886-1900.