📊 EDA Tool — Exploratory Data Analysis

A Python-based Exploratory Data Analysis tool with two interfaces:

• Terminal — interactive menu-driven CLI
• Streamlit — web-based UI with file upload support

---

📁 Project Structure

EDA Project/
│
├── app.py                    # Streamlit web application
├── main.py                    # Entry point — data loading
├── eda_utils.py               # Orchestration layer for terminal flow
├── numeric_information.py     # All numeric column analysis functions
└── categorical_information.py # All categorical column analysis functions

---

⚙️ How Each File Fits Together

main.py  ──────────────────────────────►  data_reading()
   │                                        (loads CSV)
   │  
   │    
   │
   └──► app.py  ──────────────────► numeric_information.py
        (Streamlit)              └──► categorical_information.py

---

🚀 Getting Started

Prerequisites

Install the required libraries:

pip install pandas numpy matplotlib seaborn streamlit

Running the Streamlit App

streamlit run app.py

Running the Terminal Version

python main.py

---

🌐 Streamlit App — Features (app.py)

File Upload

• Upload any .csv file directly from your desktop via the file uploader
• App halts gracefully until a file is provided

DataFrame Section

Feature	Description
Shape	Displays total rows and columns
Column Information	Shows datatype, non-null count, null count, and null percentage for every column
Sample Rows	User selects how many random rows to preview

Column Analysis Section — Numeric Columns

Feature	Description
Basic Description	Mean, Median, Std Dev, and full .describe() output
Outlier Detection	Auto-selects Z-Score method (normal distribution) or IQR method (skewed)
Missing Value Analysis	Null count, percentage, and suggested treatment
Missing Value Imputation	User chooses Yes/No to trigger imputation

Column Analysis Section — Categorical Columns

(Coming soon — plot functions pending return fig fix)

Feature	Description
Basic Summary	Unique count and mode
Value Counts	Full frequency table
Missing Value Analysis	Null count, percentage, and suggested treatment
Missing Value Imputation	User chooses Yes/No to trigger imputation

---

🖥️ Terminal App — Features (menu.py)

------------- EDA MENU -------------
1. Basic Information about Data
2. Information about a Particular Column
3. Exit
------------------------------------

For numeric columns, further options include:

1. Basic DataFrame Info
2. Describe Column
3. Skew & KDE Plot
4. Outlier Detection
5. Missing Value Calculation
6. Missing Value Imputation

---

🔧 Function Reference

numeric_information.py

Function	Description	Returns
Describing_numeric_column(df, col)	Mean, std dev, describe	mean, std_dev
calculate_skew_and_plot_kdeplot(df, col)	Skew value + KDE plot	skew_value
outlier_detection(df, col, skew, mean, std_dev)	Detects outliers using Z-Score or IQR	Q1, Q2, Q3, IQR, Lower_Bound, Upper_Bound, new_df
missing_value_calculation(df, col)	Null count, percentage, verdict	null_values, null_pct, message
whether_to_impute_missing_values(df, col, choice)	Triggers imputation based on user choice	status message
missing_value_treatment(df, col)	Applies chosen imputation method	Modifies dataframe in place

categorical_information.py

Function	Description	Returns
counting_information(df, col)	Unique count, mode, value counts	Prints to console
plot_countplot(df, col)	Bar chart or pie chart based on cardinality	Prints to console

---

🧠 Outlier Detection Logic

The outlier_detection() function automatically picks the right method based on skew:

Skew ≤ 1  →  Z-Score Method   →  Lower = mean - 3*std,  Upper = mean + 3*std
Skew > 1  →  IQR Method       →  Lower = Q1 - 1.5*IQR,  Upper = Q3 + 1.5*IQR

---

🔮 Pending / Coming Soon

• calculate_skew_and_plot_kdeplot() — needs return fig fix to render in Streamlit
• plot_countplot() — needs return fig fix to render in Streamlit
• Download button to export cleaned dataframe as CSV

---

👤 Author

Built as a learning project to practice EDA, OOP, and Streamlit development.