# Import required libraries
from sklearn.datasets import make_swiss_roll
from sklearn.decomposition import KernelPCA

# Generate synthetic data
X_swiss, t = make_swiss_roll(n_samples=1500, noise=0.3, random_state=2)

# Instantiate a KernelPCA object, specifying the kernel type 
# and the output's dimensions
pca_swiss = KernelPCA(n_components=2, kernel='sigmoid', gamma=1e-3, coef0=1, fit_inverse_transform=True)
# Transform the original data
X_pca_swiss = pca_swiss.fit_transform(X_swiss)

# Plot the original vs. reduced data