llmpic 0.2.1

LLM-powered chart generation Python SDK — create charts with natural language

llmpic

Natural Language Chart Generation Python SDK

中文（chinese）

---

from llmpic import llmPIC

lp = llmPIC(api_key="sk-...", base_url="https://api.openai.com/v1")

# Describe it. Get a chart.
lp.plot("30-day CPU usage trend").show()   # Jupyter inline display
lp.plot("CPU usage trend").save()           # → ~/llmpic_charts/

---

💡 Why llmpic?

Traditionally, creating charts in Python means wrestling with matplotlib's verbose API — plt.subplots(), ax.set_xticklabels(), fig.tight_layout() — hundreds of functions to memorize, dozens of lines for a single chart. Data scientists spend more time googling matplotlib syntax than analyzing data.

llmpic brings Python charting into the LLM era. For data scientists, analysts, quantitative researchers — anyone who needs charts from data — just describe what you want in plain language and get production-quality matplotlib charts instantly.

	Traditional matplotlib	llmpic
Lines of code	15–40 lines	1–3 lines
API knowledge	100+ functions	0 (natural language)
Chart types	Manual selection	11 types + auto-detect
Iteration	Rewrite entire block	result.edit("...")
Jupyter	plt.show() only	result.show() inline
Multi-format	Separate savefig calls	Single save()
Error recovery	Manual debugging	Auto-fix with LLM

---

✨ Features

• 🗣️ Natural Language Input — describe charts in plain English, Chinese, Japanese, or Korean
• 📊 11 Chart Types — Line, Scatter, Bar, Pie, Histogram, Heatmap, Boxplot, Area, Radar, Subplots, Auto-detect
• 📓 Jupyter Inline — result.show() renders charts directly below notebook cells
• ⚡ Async Batch — AsyncllmPIC.batch() generates multiple charts concurrently
• 🔧 Auto-Fix — failed code executions are auto-corrected by the LLM (up to 2 rounds)
• ✏️ Iterative Editing — result.edit("make bars red") refines charts with natural language
• 📦 Multi-Format — single save() with extension auto-detection (PNG/SVG/PDF), defaults to home directory
• 🌍 Multi-Language Labels — auto-detects query language and matches chart labels (zh/ja/ko/en)
• 🛡️ Dual Safety — 31 precompiled regex patterns + optional LLM semantic review
• 💻 Cross-Platform — Windows / Linux / macOS, automatic CJK font configuration

---

📦 Installation

pip install llmpic          # minimal
pip install llmpic[full]    # + pandas, seaborn, scikit-learn

Requires Python ≥ 3.10 and an OpenAI-compatible API endpoint (OpenAI, Azure, DeepSeek, etc.).

📖 Documentation

Document	Description
API Reference	Complete class, method, and parameter reference
User Guide	Advanced usage, best practices, troubleshooting
中文文档	Full documentation in Chinese
Jupyter Demos	Ready-to-run Jupyter notebooks

---

🚀 Quick Start

from llmpic import llmPIC

lp = llmPIC(
    api_key="sk-your-key",
    base_url="https://api.openai.com/v1",
    model="gpt-4o",
)

# Basic — one line, one chart
lp.plot("12-month sales trend").save("sales.png")

# With data & style
import pandas as pd
df = pd.read_csv("sales.csv")
lp.bar("Sales by region").data(df).style({
    "color_scheme": "warm",
    "figsize": [12, 7],
}).save("bar.png")

# Jupyter inline display
lp.plot("CPU usage trend").render().show()

# SVG / PDF export
lp.plot("trend").save("chart.svg")
lp.plot("trend").save("chart.pdf")

# Iterative editing
r = lp.plot("Quarterly sales: Q1=100,Q2=150").render()
r.edit("change to bar chart, red").edit("title 'Annual Report'").show()

# Default save path (no argument → home directory)
lp.plot("simple trend").save()  # → ~/llmpic_charts/chart_{timestamp}.png

Async Batch

from llmpic import AsyncllmPIC
import asyncio

async def main():
    lp = AsyncllmPIC(api_key="sk-...", base_url="https://api.openai.com/v1")

    results = await lp.batch([
        ("plot",     "12-month national sales trend"),
        ("bar",      "Sales comparison by region"),
        ("pie",      "Market share distribution"),
        ("scatter",  "Customer age vs spend"),
        ("heatmap",  "Correlation matrix"),
    ])

    for i, r in enumerate(results):
        r.save(f"chart_{i}.png")

asyncio.run(main())

---

📊 Chart Types

Method	Type
.plot()	Line
.scatter()	Scatter
.bar()	Bar
.pie()	Pie
.hist()	Histogram
.heatmap()	Heatmap
.boxplot()	Boxplot
.area()	Area
.radar()	Radar
.subplots()	Dashboard
.custom()	Auto-detect

---

📄 License

MIT © 2026 ADW-19