ETL - Volume Analyzer

View the Project on Github
README.md

Crypto Trading Volume Analysis

This repository contains a Python script that I have created to analyze the trading volume data of a cryptocurrency for my personal use.

ETL - Volume Analyzer

This project represents a comprehensive ETL (Extract, Transform, Load) pipeline, crafted in Python, intended for the analysis, clustering, and regression computation of diverse datasets. For the sake of data security and confidentiality, certain datasets, specifically Android and iOS user data, have not been disclosed.

Input Data vs Output Data of The Pipeline:

After doing feature engineering, here's the output data of the project.

Input-data Output

The ETL pipeline is structured around five primary classes:

1 - DataExtractor

Equipped with various methods, this class is responsible for extracting data from external sources.


2 - DataTransformer

This class contains a suite of methods designed to transform the data procured by the DataExtractor.


3 - DataLoader

This class encompasses several methods for efficiently loading the provided data.


4 - DataAnalyzer

Embedded with numerous methods, it facilitates Exploratory Data Analysis (EDA) on the available data.


5 - DataValidator

With several dedicated methods, it ensures data validation. Data will not proceed to analysis unless authenticated by this class.


Purposefully designed, this project aims to address pivotal questions, such as:

  • Which weekdays witness higher trading volumes?
  • Do trading patterns differ between weekends and weekdays?
  • Are there specific months that experience heightened activity?
  • How is trading volume distributed across various weekdays and months?
  • Which data clusters dominate specific months?
  • Are there clusters that are notably more frequent on certain days?
  • What is the prevailing direction of the exchange's volume trend?

Outputs:


Volume Over Time:

Vot Output

Volume with Trend Over Time:

Vtot Output

Volume Distribution:

Volume Distribution Output

Daywise Trading Volume Summary:

DWTWS Output

Volume Distribution by Day of Week:

VDBOW Output

Monthly Trading Volume Summary:

MTVS Output

Average Trading Volume by Month (Weekday vs Weekend):

wvw Output

Volume Distribution by Month:

vdbm Output

Clusters of Volume:

Clustered the days based on their volume


cov Output

Daywise Distribution of Clusters:

ddoc Output

Monthwise Distribution of Clusters:

mdoc Output

BTC Price vs Volume:

btcpricevvolume Output

VIX vs Volume:

vix Output