Data Preprocessing with Python (Manual Encoding & Imputation)

Welcome to my Data Preprocessing Project

This repository demonstrates core data preprocessing techniques in Python using Pandas, NumPy, and basic Python — all done manually without relying on any pre-built functions or libraries like get_dummies or sklearn.

It's perfect for beginners who want to understand how encoding and imputation work under the hood.

---

Project Overview

Data preprocessing is a crucial step in any Machine Learning pipeline. It ensures that the dataset is clean, consistent, and ready for modelling.

In this project, we focus on three main techniques:

1. Ordinal Encoding

   • Converting categorical data with natural order (like education levels) into numbers.
   • Example: 12th Pass = 1, Graduate = 2, Post-Graduate = 3.

2. One Hot Encoding

   • Converting categorical data with no specific order into separate binary columns (0/1).
   • Example: Cities like Delhi, Mumbai, Bangalore → City_Delhi, City_Mumbai, City_Bangalore.

3. Imputation (Mean)

   • Handling missing values (NaN) by replacing them with the mean of the column.
   • Ensures there are no gaps in the dataset for numerical analysis.

Dataset :

ID	Name	City	Education	Experience (Years)	Salary (₹)
1	Amit	Delhi	Graduate	2	32000
2	Riya	Mumbai	Post-Graduate	5	54000
3	Sam	Delhi	12th Pass	NaN	25000
4	John	Bangalore	Graduate	3	NaN
5	Neha	Mumbai	Post-Graduate	4	58000
6	Arjun	Delhi	12th Pass	1	NaN
7	Priya	Bangalore	Graduate	NaN	41000

Techniques Implemented

1. Ordinal Encoding → Mapping education levels to numeric values using basic Python.
2. One Hot Encoding → Creating city dummy columns using loops.
3. Mean Imputation → Filling missing values using manually calculated means.