import vega_datasets
import panel as pn
import pandas as pd
import altair as alt
from vega_datasets import data


pn.extension(design='bootstrap')

# we want to use vega, tell Panel to load up what we need
pn.extension('vega')


template = pn.template.BootstrapTemplate(
    title='SI649 Lab7 Viz4',
)


# Import panel and vega datasets
# load data
df1 = pd.read_csv(
    "https://raw.githubusercontent.com/dallascard/SI649_public/main/altair_hw3/approval_polllist.csv")
df2 = pd.read_csv(
    "https://raw.githubusercontent.com/dallascard/SI649_public/main/altair_hw3/approval_topline.csv")

# fix the time stamps and reorganize the data to combine approve and disapprove into one column
df2['timestamp'] = pd.to_datetime(df2['timestamp'])
df2 = pd.melt(df2, id_vars=['president', 'subgroup', 'timestamp'], value_vars=[
              'approve', 'disapprove']).rename(columns={'variable': 'choice', 'value': 'rate'})
df2.head()

# Enable Panel extensions
pn.extension()

# Define a function to create and return a plot


def create_plot(subgroup, date_range, moving_av_window):

    # Apply any required transformations to the data in pandas
    # print(pd.to_datetime(df2['timestamp']), type(date_range[0]))

    # df2_filtered = df2[(df2['subgroup'] == subgroup) & (df2['timestamp'] >= date_range[0]) & (df2['timestamp'] <= date_range[1])]
    df2_filtered = df2[(df2['subgroup'] == subgroup) & (df2['timestamp'] >= pd.Timestamp(
        date_range[0])) & (df2['timestamp'] <= pd.Timestamp(date_range[1]))]

    # Calculate the moving average
    # df2_filtered['moving_rate'] = df2_filtered['rate'].rolling(window=moving_av_window).mean()
    df2_filtered['moving_rate'] = df2_filtered['rate'].rolling(
        window=moving_av_window).mean().shift(-moving_av_window//2)
    # df2_filtered.loc[:, 'moving_rate'] = df2_filtered['rate'].rolling(window=moving_av_window).mean()
#
    # Line chart for moving average with df2_filtered
    line = alt.Chart(df2_filtered).mark_line().encode(
        x='timestamp:T',
        y='moving_rate:Q',
        color=alt.value('red')
    ).transform_filter(
        # keep only approved polls
        alt.FieldOneOfPredicate(field='choice', oneOf=['approve'])
    )

    # Scatter plot with individual polls
    scatter = alt.Chart(df2_filtered).mark_circle().encode(
        x='timestamp:T',
        y='rate:Q',
        color=alt.value('gray'),
        opacity=alt.value(0.7),
        tooltip=['timestamp:T', 'rate:Q'],
        size=alt.value(10)
    ).transform_filter(
        # keep only approved polls
        alt.FieldOneOfPredicate(field='choice', oneOf=['approve'])
    )

    # Put them together
    plot = scatter + line
    # change y axis range of the plot to 30 to 60

    # plot = plot + line
    plot = plot.properties(title="Approval Ratings for Joe Biden")
    plot = plot.encode(
        y=alt.Y('rate:Q', scale=alt.Scale(domain=[30, 60]))
    )
    # Return the combined chart
    return plot


# Create the selection widget using subgroup column
selection_subgroup = pn.widgets.Select(
    name='Subgroup', options=df2.subgroup.unique().tolist())

# Create the slider for the date range
df2['timestamp'] = pd.to_datetime(df2['timestamp'])
date_range_slider = pn.widgets.DateRangeSlider(name='Date Range', start=df2.timestamp.min(
), end=df2.timestamp.max(), value=(df2.timestamp.min(), df2.timestamp.max()))

# Create the slider for the moving average window
moving_av_window_slider = pn.widgets.IntSlider(
    name='Moving Average Window', start=1, end=30, step=1, value=1)
# Bind the widgets to the create_plot function
create_plot_wgt = pn.bind(create_plot, subgroup=selection_subgroup,
                          date_range=date_range_slider, moving_av_window=moving_av_window_slider)

# Combine everything in a Panel Column to create an app
main_col = pn.Column(
    create_plot_wgt,
    selection_subgroup,
    date_range_slider,
    moving_av_window_slider,
)
# set the app to be servable
main_col.servable()