from vega_datasets import data
from scipy import stats
from bokeh.plotting import figure
from bokeh.models import ColumnDataSource
import panel as pn
import numpy as np

# Tell banel to use Bokeh
pn.extension()

# Get the data
source = data.movies()
temp = sorted(source['IMDB_Rating'].dropna().values)

# compute an initial number of bins
n_int_bins = int(np.ceil(max(temp))+ 1 - np.floor(min(temp)))

# define a function that takes the parameters and creates the plot
def create_plot(bandwidth=1.0, bins=n_int_bins):
    plot = figure(width=300, height=300, toolbar_location=None)
    
    # Compute the histogram with the specified number of bins
    hist, edges = np.histogram(temp, bins=bins)

    # normalize and plot it
    hist = hist / hist.sum()
    quad = plot.quad(top=hist, bottom=0, left=edges[:-1], right=edges[1:],
             fill_color="grey", line_color="white", alpha=0.5)

    # Compute the density with the specified bandwidth
    kernel = stats.gaussian_kde(temp, bw_method=bandwidth)
    x = np.linspace(min(temp), max(temp), 100)
    y = kernel(x)
    col_source = ColumnDataSource(data=dict(x=x, y=y))
    line = plot.line('x', 'y', source=col_source, alpha=1.0, width=2)

    return plot

# Create widgets for the bandwidth and number of bins
bw_widget = pn.widgets.FloatSlider(name="Bandwidth", value=1.0, start=0.03, end=2.0, step=0.02)
bins_widget = pn.widgets.IntSlider(name="Number of Bins", value=n_int_bins, start=1, end=n_int_bins*10)

# bind the sliders to the plotting function
bound_plot = pn.bind(create_plot, bandwidth=bw_widget, bins=bins_widget)

# Combine everything together
first_app = pn.Column(bw_widget, bins_widget, bound_plot)
first_app.servable()