Exploratory Data Analysis (EDA) using Python

Overview

This project demonstrates skills applied in a technical assessment for a data analytics role.
It focuses on analyzing and tagging a dataset of customer complaints and repair information to derive actionable insights.

Key objectives:

• Clean and preprocess raw data.
• Tag fields such as Root Cause, Symptom Condition, Symptom Component, Fix Condition, and Fix Component.
• Identify critical columns and patterns for stakeholders.
• Generate visualizations to communicate insights effectively.

---

Dataset

The dataset contains detailed information on customer complaints and repairs, including:

• Vehicle & Transaction Info: VIN, TRANSACTION_ID, REPAIR_DATE, VEH_TEST_GRP, COUNTRY_SALE_ISO, ORD_SELLING_SRC_CD
• Customer Complaints: CUSTOMER_VERBATIM, CORRECTION_VERBATIM, COMPLAINT_CD_CSI, COMPLAINT_CD
• Parts & Labor: CAUSAL_PART_NM, GLOBAL_LABOR_CODE_DESCRIPTION, TRANSACTION_CATEGORY, REPORTING_COST, TOTALCOST, LBRCOST
• Vehicle Specs: PLATFORM, BODY_STYLE, ENGINE, TRANSMISSION, ENGINE_DESC, TRANSMISSION_DESC, ENGINE_SOURCE_PLANT, TRANSMISSION_SOURCE_PLANT
• Dealership Info: LAST_KNOWN_DLR_NAME, LAST_KNOWN_DLR_CITY, REPAIRING_DEALER_CODE, DEALER_NAME, REPAIR_DLR_CITY, STATE, DEALER_REGION, REPAIR_DLR_POSTAL_CD
• Repair Metadata: REPAIR_AGE, KM, MEDIA_FLAG, VIN_MODL_DESGTR, LINE_SERIES, LAST_KNOWN_DELVRY_TYPE_CD, NON_CAUSAL_PART_QTY, SALES_REGION_CODE

Note: Sensitive information has been anonymized for privacy.

---

Approach

1. Data Cleaning

   • Removed duplicates and null values.
   • Standardized inconsistent entries for parts, labor codes, and repair notes.

2. Data Tagging

   • Applied structured logic to tag root causes, symptoms, and fixes.
   • Ensured consistency across multiple components and conditions.

3. Analysis & Visualization

   • Selected top 5 critical columns for stakeholder insights.
   • Created visualizations (bar plots, distribution charts) to highlight key patterns.
   • Derived actionable insights to support data-driven decision making.

---

Key Insights

• Recurring complaints help prioritize preventive actions.
• Symptom and fix patterns indicate areas for process improvement.
• Data-driven analysis can guide repair teams to reduce repeat issues.
• Certain parts and modules frequently cause customer-reported problems.

---

Tools & Technologies

• Python 3.x
• Pandas, NumPy
• Matplotlib, Seaborn
• Jupyter Notebook

---

Conclusion

This repository serves as a hands-on example for aspiring Python data analysts to understand how to:

• Load and clean real-world datasets.
• Perform exploratory data analysis (EDA) to extract meaningful insights.
• Identify key features that matter to stakeholders.
• Create effective visualizations to communicate findings.

By studying this repo, learners can gain practical experience and ideas for structuring their own data analysis projects in Python.

Happy learning ❤️