Manuales
Aquí serán incluidos manuales, anotaciones y demás que vaya escribiendo. Todo lo que sean articulos ó documentanciones que considere interesanes las iré listando aquí por temas.
DATA HANDLING
Data format
# save data as feather file format
df.to_feather('filename.feather')
# read feather file
df1 = pd.read_feather('filename.feather')
Missing values imputation
- sklearn - Imputation of missing values.
- Univariate feature imputation.
- Multivariate feature imputation.
- Nearest neighbors imputation.
- Marking imputed values.
Data discretization: continuous to discrete values
- sklearn - sklearn.preprocessing.KBinsDiscretizer: Bin continuous data into intervals. It is available three possible strategies: ‘uniform’, ‘quantile’, ‘kmeans’.