NJmatML 0.0.4

Automates machine learning model construction for materials science via NJmatML. Please check https://github.com/Zhang-NJ-Lab/NJmatML-Functions/blob/main/define.ipynb

定义了一些函数用于机器学习建模，数据使用的是表面修饰钙钛矿薄膜的水环境光电压. The test dataset is photocurrent voltages for surface modified halide perovksite films in water (or formation energy). file_name('train1.csv') hist() heatmap_before() feature_select(23,0) #第一个为剩的特征个数，第二个一般都为0 heatmap_afterRFE() FeatureImportance_before(80,8,10,4) #rotation=80, fontsize=8, figure_size_xaxis=10,figure_size_yaxis=4 FeatureImportance_afterRFE(80,12,5,4) #rotation=80, fontsize=12, figure_size_xaxis=5,figure_size_yaxis=4 xgboost_default() xgboost_modify(1000,200,0.2,0,0.9,0.8,0.2)