Get Started¶

How to install Sketchkit¶

Install Anaconda¶

Anaconda is a free, open-source distribution of Python and R designed for data science, machine learning, and scientific computing. It includes many popular libraries and tools, and comes with the Conda package manager to easily manage environments and deployments.

Download Anaconda¶

Download the Anaconda installer that matches your operating system from one of these sources:

Official: Anaconda Downloads (Global)
Mirror: Tsinghua Mirror (Faster in China)

Installation Steps¶

Run the installer and follow the setup wizard
Important: During installation, check the box “Add Anaconda to my PATH environment variable” so you can use the conda command directly in the terminal
Restart your terminal after installation completes

Verify Installation¶

Open terminal/PowerShell and type:

conda --version

If you see the version information (e.g., conda 24.7.1), then Anaconda is installed successfully.

Troubleshooting:

If conda command is not found, restart your terminal or manually add Anaconda to your PATH
On Windows, try using “Anaconda Prompt” from the Start menu

Install SketchKit¶

Configure API keys¶

The API keys are stored in ~/sketchkit/config. The keys should be stored like the following:

OPENAI_API_KEY="sk-proj-xxxxxxxxxxxxx"

Step 1: Prepare SketchKit Files¶

Extract SketchKit.zip to a folder named SketchKit

Step 2: Open Terminal¶

Navigate to the SketchKit folder in your terminal:

Method 1 (File Manager):

Right-click in the SketchKit folder and select “Open in Terminal” or “Open PowerShell here”

Method 2 (Command Line):

cd /path/to/SketchKit

Replace /path/to/SketchKit with the actual path to your SketchKit folder.

Step 3: Install SketchKit¶

Run the following commands:

# Create a new conda environment with Python 3.12
conda create -n sketchkit python==3.12

# Activate the environment  
conda activate sketchkit

# Install SketchKit in development mode
pip install -e .

After installation, you can import sketchkit in your python program.

Step 4: Verify Installation¶

Test your installation:

import sketchkit
from sketchkit.datasets import QuickDraw
print("SketchKit installed successfully!")

Note: Use pip install -e . (with -e) for development installation, which allows you to modify the code and see changes immediately.

Example¶

Load dataset¶

SketchKit provides several options for loading datasets:

Basic Loading (Recommended for beginners)¶

# Load metadata only (fast, low memory usage)
dataset = QuickDraw()

Load All Data to Memory

# Load all sketches to memory (faster access, high memory usage)
dataset = QuickDraw(load_all=True)

Use Alternative Download Source

# Download from CISLab mirror (may be faster in some regions)
dataset = QuickDraw(cislab_source=True)

Working with Dataset Metadata¶

All datasets provide a items_metadata attribute that helps you search and filter sketches:

from sketchkit.datasets import QuickDraw

# default load
dataset = QuickDraw()

# search data with "category = cat" and "split = train"
cats = dataset.items_metadata[
    (dataset.items_metadata["category"] == "cat")
    & (dataset.items_metadata["split"] == "train")
]

# get sketches from the dataset
cats_sketch = [dataset[row.id] for _, row in cats[:100].iterrows()]

Render the sketch¶

Once you have loaded sketches, you can render them as raster images using SketchKit’s CairoRenderer. This converts vector sketch data into images (e.g., PNG).

from PIL import Image
from sketchkit.renderer.cairo_renderer import CairoRenderer

# Load a sketch from the dataset
sketch = dataset[100]

# Create a renderer with canvas size and background color (RGB)
renderer = CairoRenderer(800, (1, 1, 1))  # 800x800 canvas, white background
raster_image = renderer.render(sketch)

# Save the rendered image
outpath = "my_sketch.png"
raster_image_png = Image.fromarray(raster_image, "RGB")
raster_image_png.save(outpath, "PNG")
print(f"Sketch saved to {outpath}")