SelectSampler package

This package allows you to fetch the dataset samples from the holoviews scatterplot as a dataframe. It only works in Jupyter notebook environment. An example usage is the following:

import numpy as np
sampler = ScatterSampler()
# Arbitrary datasets for demonstration
mat = np.random.randint(0, 100, (100, 2))
df = pd.DataFrame(mat, columns=['x', 'y'])
sampler.setup_holoviews()
sampler.setup_scatter(df, 'x', 'y')
scatter_chart = sampler.show_scatter_chart()
scatter_chart

sampler.extract_df()

Steps:

1. Run sampler.setup_holoviews() first to ensure you have the bokeh extension installed.
2. Then run sampler.setup_scatter(df, 'x', 'y') to setup the scatter plot.
3. Run sampler.show_scatter_chart() to display the scatter plot. You can select the samples on the sidebars of the scatter plot.
4. Finally, run sampler.extract_df() to get the selected samples in a dataframe (you need to run this in the separate cell).