a pipeline for quick data analysis and machine learning
file_operate:
123
- read file and do base operate like auto asdtype etc.
- it support multi type files operates together.
preprocess:
create a preprocessor to preprocessing(dropna,fillna,dropoutlyers...)
you can select the file and cols by passing dict-type args.
analysis:
base on the previous manipulations,we get clean datas,we can now acutally start the analysis tasks:
correlation:
get correlations between value-type features and labels.
compare the correlation between class-type features and labels.
ploting:
draw graceful basic figures(linear,scatter,bar) fast and scalably.
modeling:
we provide base ml models to complete classification or regression tasks
listing:
gbdt:xgboost,light gbm,radom forests
norm:svc,linear,logistic,bayes
timesequence:ARMA,ARIMA
nn:
DeepLearning:\
statistic_test: normality test correlation test significance test parametric test nonparametric test
