🐼 PandaScrub: Interactive Data Studio & Python Visualizer

The reactive 2-way visual studio for Pandas, DataFrames & Python in Visual Studio Code.
Code in standard Python, press Ctrl + S, and watch your DataFrames, Charts, and Pivot Tables update live beside your code.

⚡ 30-Second Quick Start

Get up and running in 3 simple steps:

Step 1: Open Your Dataset in PandaScrub

In VS Code's file explorer, right-click any .csv, .tsv, .xlsx, or .parquet file.
Select "Open with PandaScrub Data Studio". (Or right-click $\to$ Open With... $\to$ PandaScrub Data Studio).

Step 2: Open Your Python File Side-by-Side

Open or create a Python file (e.g. analysis.py) in a split editor tab beside PandaScrub.

Paste this starter code:

import pandas as pd

# Load dataset
df = pd.read_csv('titanic.csv')

# Filter passengers
df = df[df['Age'] > 30]

Step 3: Hit `Ctrl + S` & Watch the Magic!

The moment you save analysis.py with Ctrl + S, PandaScrub instantly extracts your DataFrame and renders it live on screen with full sorting, filtering, and column profiling!

🚀 The 4 Core Superpowers

1. 📊 Interactive Data Table & Multi-Variable Switcher

Any Variable Name: PandaScrub detects DataFrames with any name (df, customers, sales, titanic, x, pivot_df).
Variable Switcher Dropdown: Have multiple DataFrames or pd.Series in your script? PandaScrub lists all of them in the top-left dropdown (Var: df (891×12)) so you can switch and inspect them with 1 click.
Instant Column Profiler: View data types (int, float, string, date), unique counts, null percentages, and data quality scores at a glance.
60 FPS Virtual Grid: Silky-smooth scrolling through 100,000+ rows with zero lag.

2. 🧹 1-Click Visual Data Cleaning (Recipe Engine)

Drop Nulls & Impute: Remove missing values or fill them with Mean, Median, Mode, or custom constants.
Text & String Cleaning: Trim whitespace, convert to lowercase/UPPERCASE/Title Case, and strip special characters.
Type Conversion & Slicing: Cast columns to numeric, parse messy dates, and filter rows visually.
⚡ Non-Destructive Code Injection: Click Insert Pipeline at any time. PandaScrub automatically writes clean, reproducible PEP-8 Pandas code directly into your Python script!
```
# >>> PANDASCRUB MANAGED PIPELINE (START) >>>
# Step 1: Drop Missing Values
df = df.dropna()

# Step 2: Trim Whitespace
df['Name'] = df['Name'].str.strip()
# <<< PANDASCRUB MANAGED PIPELINE (END) <<<
```
(Your imports, custom functions, and surrounding logic remain 100% untouched. Press Ctrl + Z in your editor to undo anytime!)

3. 🔄 Visual Pivot Table & GroupBy Studio

Cross-Tabulations in Seconds: Drag and select your Rows (Index), Columns (Pivot), and Value columns with aggregations (sum, mean, count, median, min, max, std).
Heatmap Gradient Shading: Visually spot outliers and high-value clusters with automatic cell coloring.
Flat GroupBy Support: Automatically collapses into standard 1D df.groupby() when no column dimension is selected.
⚡ Insert into Python Script: Click the green Insert into Python Script button to generate production pd.pivot_table(df, ...) code directly into your script!

4. 📈 Matplotlib & Seaborn Chart Playground

Live Side-by-Side Plotting: Write free-hand Matplotlib / Seaborn code and preview publication-grade charts immediately.
Instant Presets: 1-click starter templates for Bar Charts, Line Trends, Scatter Plots, Box Plots, KDE Density Curves, and Correlation Heatmaps.
Flexible Layouts: Switch between side-by-side columns, stacked rows, code-focus, or canvas-focus modes.
1-Click Export: Download high-res PNGs or click Insert into Script to inject your plotting code into your project.

💡 Practical Python Examples to Try

Example A: Inspect Trailing Expressions

PandaScrub automatically handles trailing .describe(), .info(), and .head():

import pandas as pd
df = pd.read_csv('titanic.csv')

# Instantly displays an 8-row statistical summary table in the GUI:
df.describe()

Example B: Series Value Counts

import pandas as pd
df = pd.read_csv('titanic.csv')

# Visualized as a clean 2-column table with category and frequency counts:
df['Embarked'].value_counts()

Example C: Cross-Tabulation Pivot

import pandas as pd
df = pd.read_csv('titanic.csv')

# Generates a summarized matrix of average fare by Class and Gender:
pivot_df = pd.pivot_table(
    df,
    index='Pclass',
    columns='Sex',
    values='Fare',
    aggfunc='mean'
).round(2)

❓ Frequently Asked Questions (FAQ)

Q: Do I have to name my DataFrame `df`?

A: No! You can name it whatever you like: customers, raw_data, sales_2026, x, etc. PandaScrub inspects Python's AST and finds your DataFrames automatically.

Q: Which Python file does PandaScrub connect to?

A: PandaScrub connects to whichever .py file you have open beside it or whichever file you just saved (Ctrl + S). If no Python file is open, it will offer to create analysis.py for you.

Q: Will PandaScrub modify my Python code without my permission?

A: Never. Code is only written into your script when you explicitly click an ⚡ Insert button. All injected code is safely contained inside clearly marked # >>> PANDASCRUB MANAGED PIPELINE >>> boundaries.

Q: How heavy is this extension?

A: PandaScrub is ultra-lightweight at under 400 KB (380 KB). It has zero heavy binary dependencies and delegates execution directly to your local Python interpreter.

🛡️ License

MIT License • Built with ❤️ for Data Analysts, Scientists, and Python Engineers.

PandaScrub: Interactive Data Studio & Python Visualizer

PandaScrub