|
import numpy as np |
|
import matplotlib.pyplot as plt |
|
from sklearn import neighbors |
|
import gradio as gr |
|
|
|
def train_and_plot(weights, n_neighbors): |
|
np.random.seed(0) |
|
X = np.sort(5 * np.random.rand(40, 1), axis=0) |
|
T = np.linspace(0, 5, 500)[:, np.newaxis] |
|
y = np.sin(X).ravel() |
|
|
|
|
|
y[::5] += 1 * (0.5 - np.random.rand(8)) |
|
|
|
knn = neighbors.KNeighborsRegressor(n_neighbors, weights=weights) |
|
fit = knn.fit(X, y) |
|
y_ = knn.predict(T) |
|
score = knn.score(T, y_) |
|
|
|
plt.figure() |
|
plt.scatter(X, y, color="darkorange", label="data") |
|
plt.plot(T, y_, color="navy", label="prediction") |
|
plt.axis("tight") |
|
plt.legend() |
|
plt.title("KNeighborsRegressor (k = %i, weights = '%s')" % (n_neighbors, weights)) |
|
|
|
plt.tight_layout() |
|
return plt, score |
|
|
|
|
|
with gr.Blocks() as demo: |
|
link = "https://scikit-learn.org/stable/auto_examples/neighbors/plot_regression.html#sphx-glr-auto-examples-neighbors-plot-regression-py" |
|
gr.Markdown("## Nearest Neighbors Regression") |
|
gr.Markdown(f"This demo is based on this [scikit-learn example]({link}).") |
|
gr.HTML("<hr>") |
|
gr.Markdown("In this demo, we learn a noise-infused sine function using k-Nearest Neighbor and observe how the function learned varies as we change the following hyperparameters:") |
|
gr.Markdown("""1. Weight function |
|
2. Number of neighbors""") |
|
|
|
with gr.Row(): |
|
weights = gr.Radio(['uniform', "distance"], label="Weights", info="Choose the weight function") |
|
n_neighbors = gr.Slider(label="Neighbors", info="Choose the number of neighbors", minimum =1, maximum=15, step=1) |
|
|
|
|
|
with gr.Row(): |
|
with gr.Column(scale=2): |
|
plot = gr.Plot(label="KNeighborsRegressor Plot") |
|
with gr.Column(scale=1): |
|
num = gr.Textbox(label="Test Accuracy") |
|
|
|
|
|
weights.change(train_and_plot, inputs=[weights, n_neighbors], outputs=[plot, num]) |
|
n_neighbors.change(train_and_plot, inputs=[weights, n_neighbors], outputs=[plot, num]) |
|
|
|
|
|
if __name__ == "__main__": |
|
demo.launch() |
