DataAdvisor (data-advisor)

Autonomous Dataset Auditor, Diagnostic Engine & IoT Exam Code Bank
Engineered for students, practitioners, and engineers during timed practical examinations and data analysis tasks.
100% Offline: Requires zero internet connection to inspect datasets, search code, or retrieve solutions.

⚡ Quickstart

1. Search Any Keyword Across the Code Bank

Search instantly across all IoT questions, Arduino sketches, and Python scripts:

# Terminal CLI
data-advisor --search "dht22"
data-advisor --search "coap"
data-advisor --search "mqtt"
data-advisor --search "random forest"

Or inside a Python script / Jupyter Notebook:

from data_advisor import search

# List all available questions and experiments:
search()

# Search for any topic or keyword:
search("dht22")
search("coap")
search(["esp32", "mqtt"])

2. View / Retrieve Complete Question & Code Files (get_file)

Retrieve the full question statement, README setup instructions, and all source codes:

# Terminal CLI
data-advisor --get Q1
data-advisor --get Q3
data-advisor --get 08
data-advisor --get 15 --line-numbers

Or inside Python:

from data_advisor import get_file, read_file

# Print the complete question, instructions, and code:
get_file("Q1")
get_file("Q3", line_numbers=True)

# Or return the content as a string:
content = read_file("08")

3. Audit Any Exam Dataset & Generate Python Code (data_advisor)

When given an unknown dataset (CSV, TSV, Excel, Parquet):

# Terminal CLI: Print diagnostic report + executable Python solution:
data-advisor exam_dataset.csv --code

# Specify target column and save generated Python script to a file:
data-advisor exam_dataset.csv --target churn --output solution.py

Or inside Python:

from data_advisor import data_advisor

# 1-Line: Outputs diagnostic report AND prints full executable Python code:
data_advisor("exam_dataset.csv")

# With target variable and script export:
data_advisor("exam_dataset.csv", target="churn", output_script="solution.py")

📦 IoT Code Bank Architecture (21 Questions)

The code_bank/ directory contains 21 comprehensive IoT questions stored as structured .txt files:

Structure of Each Question File:

• Title Banner: Question / Experiment header.
• FULL QUESTION: Complete official question text with marks and evaluation rubrics.
• SHORT QUESTION: Concise 1–2 sentence summary / objective.
• README & SETUP INSTRUCTIONS: Hardware wiring, pinout diagrams, dependencies, execution commands, and expected outputs.
• SOURCE CODE FILES: Complete executable code files clearly numbered (e.g., 1. server.py, 2. client.py, 3. app.py or 1. sketch.ino).

Complete List of 21 Questions:

Adding More Questions: You can add your own .txt files to code_bank/ at any time. The search and retrieval functions automatically detect new files dynamically.

🔒 100% Offline Capability

• Zero Network Required: Does not connect to the internet to run searches, load code files, or inspect datasets.
• Local Fallback Datasets: The machine learning scripts support offline CSV datasets (city_day.csv, historical_weather_kochi.csv, datatraining.txt).
• Local Protocol Testing: CoAP server/client examples bind to 127.0.0.1:5683 (localhost UDP) and local MQTT uses localhost:1883.

🚀 Building & Publishing to PyPI

1. Build the Package

python -m build

(Builds the universal wheel and source distribution bundling the complete library and code bank).

2. Upload to PyPI

python -m twine upload dist/*

• Username: __token__
• Password: <paste your PyPI API token starting with pypi-...>

💻 Installation on Other Machines

On any lab system, exam machine, or virtual environment:

# Install directly from PyPI
pip install data-advisor

# Or install locally in editable mode
pip install -e .

CLI Command Summary: