Wheels, Utilities, Tools of data science
Now it contains following modules:
Design for feature engineerings and preprocessing of data science, including
- binary_search
- bucketized_column
- discrete_for_lookupTable
- cross_column
- sparse_column
Methods of analysing and visualizing log file, including
- analyse_bigdl : to analyse bigdl log file to get some visualization of training process.
- analyse_pytorch : to analyse pytorch log file to get some visualization of training process.
Data Transforms and preprocessing of images, text, or records etc, including
- Normalize : Normalize(tensor, mean, std)
- generate_data : generate a batch of data