sabia-utils 0.1.6

Group of utilities for Sabia

Sabia Utils

This is a collection of utilities for Sabia.

Concat Module

This module is used to concatenate files.

Concatenate all files from a path

• Returns a concatenated dataframe from all files in a path.
• Can save the concatenated dataframe to a file.

from sabia_utils import concat

concat.concatenate_all_from_path(
    path='path//to//files',
    output_file='output//file//path', # optional
    fine_name='file_name'           # optional 
)

Concatenate some files from a path

• Returns a concatenated dataframe from some files in a path.
• Can save the concatenated dataframe to a file.

from sabia_utils import concat

concat.concatenate_files(
    path='path//to//files',
    files=['file1', 'file2'],
    output_file='output//file//path', # optional
    fine_name='file_name'           # optional 
)

Copy files from a path to another

• Verify if the files exist in the path before copy.

from sabia_utils import group

group.copy_new_files(
    PATH_IN='path//to//files1',
    PATH_OUT='path//to//files2'
)

Group Module

This module is used to group files.

Process files in both paths

• Verify if the files exist in the path before process.

from sabia_utils import group

group.process_existent_files(
    PATH_IN='path//to//files1',
    PATH_OUT='path//to//files2'
)

Process all files

• apply the function of copy and process files between the paths.

from sabia_utils import group

group.process_all_files(
    PATH_IN='path//to//files1',
    PATH_OUT='path//to//files2'
)

Pre_process Module

This module is used to pre_process parquet files.

Process all parquet files

• Define a class that inherit sabia_utils.pre_process.Processing

• Override method apply_to_df(self, df, column), defining the pre-processing to be applied

• Apply this function on a folder containing parquet files to process them.

from sabia_utils.pre_process import Processing
from sabia_utils import pre_process

class MyProcessor(Processing):
    def apply_to_df(self, df, column): 
        # Your pre-processing steps

pre_process.pre_process_parquets(
    folder_path='path//to//folder',
    colomun_to_pre_process='column_name_to_be_processed',
    pre_processed_column='processed_column_name',
    processor=MyProcessor()
)