an elegant datasets factory

• Free software: MIT license
• Documentation: https://rawbuilder.readthedocs.io.

• Schema oriented datasets builder

Terminal:

# Import the package into any python app
import rawbuilder as rw

# Init the dataset object as ds
ds = rw.DataSet(
    size=1000,
    task='user',
)

# Build the dataset
ds.build()

# Optionals
ds = rw.DataSet(
    size=1000,
    task='user',
    schema_path='where/to/read/schema/from',
    schema_dict='{'user':{'id':'int'}}'
)

df = ds.build(
    output_path='your/output/directory',
    export_csv=True,
    return_df=True
)

• The Schema is a JSON object that describes three main components.
• The model names, the column names, and the data types per column.
• Note the below code-block, The model name is "Student", and it contain 4 columns [id,first_name,email,math_test_results].
• Each property of the model "student" is called a task and it has its columns and data description.

Student data model example:

"student": {
    "id": "int",
    "first_name": "first_name",
    "last_name": "last_name",
    "email": "email",
    "math_test_results": "random_int between,0,30"
}

• int: build a column of integers between 1 and requested dataset size.
• decrement: build a column of decremented integers between the requested size and 1.
• random_int: build a column of random integers between 0 and 100 by default.
• random_float: build a column of random floats between 0 and 1 by default.
• first_name: build a column of first names.
• last_name: build a column of last names.
• email: build a column of fake emails.
• password: build a random string passwords with default length of 12 characters.

Combine Data Modifiers to the above data types, it can adjust values, change the data nature, and gives more control over the final output.

Modifiers syntax is simple:

"modifier,argument_1,arg_2,arg_*"

Use the modifier between to generate random integer column between 0 and 30:

"math_test_results": "random_int between,0,30"

Use this modifier to set the high-end and low-end for a specific data type

Syntax:

"between,10,1000"

Supported with

random_int:

"math_test_results": "random_int between,0,30"

random_float:

"heights": "random_float between,1.30,1.80"

password:

"password": "password between,12,12"