Quick Start Guide

This guide will help you get started with SheetWise in just a few minutes.

Basic Example

Here’s a simple example to compress and encode a spreadsheet:

import pandas as pd
from sheetwise import SpreadsheetLLM

# Initialize
sllm = SpreadsheetLLM()

# Load your spreadsheet
df = pd.read_excel("data.xlsx")

# Compress and encode
encoded = sllm.compress_and_encode_for_llm(df)
print(encoded)

Auto-Configuration

Let SheetWise automatically optimize compression settings:

from sheetwise import SpreadsheetLLM

sllm = SpreadsheetLLM()
df = pd.read_excel("data.xlsx")

# Auto-configure based on spreadsheet characteristics
auto_compressed = sllm.compress_with_auto_config(df)

Token Budget Control

Ensure your output fits within a token limit:

from sheetwise import SpreadsheetLLM

sllm = SpreadsheetLLM()
df = pd.read_excel("large_file.xlsx")

# Automatically adjust compression to fit 4000 tokens
encoded = sllm.encode_to_token_limit(df, max_tokens=4000)

Working with Formulas

Extract and analyze Excel formulas:

from sheetwise import FormulaParser

parser = FormulaParser()
formulas = parser.extract_formulas("workbook.xlsx")

# Build dependency graph
parser.build_dependency_graph()

# Get formula impact
impact = parser.get_formula_impact("Sheet1!A1")
print(impact)

Advanced Table Detection

Detect and classify tables in your spreadsheet:

from sheetwise import SmartTableDetector

detector = SmartTableDetector(header_detection=True)
tables = detector.detect_tables(df)

for table in tables:
    print(f"Table type: {table.table_type}")
    print(f"Has headers: {table.has_headers}")
    print(f"Header rows: {table.header_rows}")
    print(f"Header cols: {table.header_cols}")

Multi-Sheet Workbooks

Process entire workbooks:

from sheetwise import WorkbookManager, SpreadsheetLLM

workbook = WorkbookManager()
sheets = workbook.load_workbook("workbook.xlsx")

# Detect cross-sheet references
refs = workbook.detect_cross_sheet_references()

# Compress entire workbook
sllm = SpreadsheetLLM()
compressed = workbook.compress_workbook(sllm.compressor)
encoded = workbook.encode_workbook_for_llm(compressed)

Visualization

Generate visual reports:

from sheetwise import CompressionVisualizer, SpreadsheetLLM

sllm = SpreadsheetLLM()
visualizer = CompressionVisualizer()

df = pd.read_excel("data.xlsx")
compressed = sllm.compress_spreadsheet(df)

# Create heatmap
fig = visualizer.create_data_density_heatmap(df)
fig.savefig("heatmap.png")

# Generate HTML report
html = visualizer.generate_html_report(df, compressed)
with open("report.html", "w") as f:
    f.write(html)

Command Line Interface

Use SheetWise from the command line:

# Basic usage
sheetwise input.xlsx -o output.txt

# With auto-configuration
sheetwise input.xlsx --auto-config --verbose

# Run demo
sheetwise --demo

# JSON output
sheetwise input.xlsx --format json

Next Steps