SHINY PYTHON

Interactive visualisations

UI vs SERVER

# Here goes everything that should appear in the app and the user can interact with
app_ui = ui.page_fluid(
    # Add title
    "Emissions Aotearoa",
    # Add selection of industry to display
    #ui.input_radio_buttons("industry", "Select industry", sectors),
    ui.input_checkbox_group("industry_check", "industry", sectors),
    ui.output_plot("emission_plot"),
    # Add table
    ui.output_table("emissions_table"),
)


# Here goes everything that happens in the background: data, computations, etc...
def server(input, output, session):
    # Load dataset
    infile = Path(__file__).parent / "emissions_data_simple.csv"
    df = pd.read_csv(infile)

Preliminaries

from shiny import App, render, ui, reactive

# You can import the libraries that you need
from pathlib import Path
import pandas as pd
from matplotlib import pyplot as plt

UI - Radio buttons

# Define list/dictionary with possible sectors
sectors = ["Agriculture", "Households", "Road transport"]

# Here goes everything that should appear in the app and the user can interact with
app_ui = ui.page_fluid(
    # Add title
    "Emissions Aotearoa",
    # Add selection of industry to display
    ui.input_radio_buttons("industry_check", "Select industry", sectors)
)

Then run the app

Server - Load data

def server(input, output, session):
    # Load dataset
    
    @output
    @render.table
    def emissions_table():
      infile = Path(__file__).parent / "emissions_data_simple.csv"
      df = pd.read_csv(infile)
      return df

app = App(app_ui, server)

UI - show data

# Define list/dictionary with possible sectors
sectors = ["Agriculture", "Households", "Road transport"]

# Here goes everything that should appear in the app and the user can interact with
app_ui = ui.page_fluid(
    # Add title
    "Emissions Aotearoa",
    # Add selection of industry to display
    ui.input_radio_buttons("industry_check", "Select industry", sectors)
)

UI - show data

# Define list/dictionary with possible sectors
sectors = ["Agriculture", "Households", "Road transport"]

# Here goes everything that should appear in the app and the user can interact with
app_ui = ui.page_fluid(
    # Add title
    "Emissions Aotearoa",
    # Add selection of industry to display
    ui.input_radio_buttons("industry_check", "Select industry", sectors),
    # Add table
    ui.output_table("emissions_table")
)

Ractivity - How to make things change

Move import outside the table-creating function

# Load dataset
    infile = Path(__file__).parent / "emissions_data_simple.csv"
    df = pd.read_csv(infile)

Create a reactive element

# Load dataset
    infile = Path(__file__).parent / "emissions_data_simple.csv"
    df = pd.read_csv(infile)
# Create a reactive table for dysplaying and plotting
    df_table = reactive.Value(df)

Ractivity - How to make things change

Create a “nameless” function to update the table

    # Update the table when checkboxes are selected/unselected
    @reactive.Effect
    @reactive.event(input.industry_check)
    def _():
        # Always re-set the table to the full dataset, otherwise we break the dataframe
        df_table.set(df)
        new_val = df_table()[df_table()["anzsic_descriptor"] == input.industry_check()]     # Use this for radio buttons
        df_table.set(new_val)

Ractivity - How to make things change

Create a “nameless” function to update the table

    # Update the table when checkboxes are selected/unselected
    @reactive.Effect
    @reactive.event(input.industry_check)
    def _():
        # Always re-set the table to the full dataset, otherwise we break the dataframe
        df_table.set(df)
        new_val = df_table()[df_table()["anzsic_descriptor"] == input.industry_check()]     # Use this for radio buttons
        df_table.set(new_val)
    
    # Define table
    @output
    @render.table
    def emissions_table():
        return df_table()

Server - Add a plot

Add the following lines to the server function

# Define plot
    @output
    @render.plot
    def emission_plot():
        fig, ax = plt.subplots()
        for key, grp in df_table().groupby("anzsic_descriptor", group_keys=True):
            ax = grp.plot(ax=ax, kind='line', x="year", y="data_value", label=key, marker='o')

UI - Add a plot

Add the following line to the UI function

# Here goes everything that should appear in the app and the user can interact with
app_ui = ui.page_fluid(
    # Add title
    "Emissions Aotearoa",
    # Add selection of industry to display
    ui.input_radio_buttons("industry_check", "Select industry", sectors),
    # Add table
    ui.output_table("emissions_table"),
    # Add plot
    ui.output_plot("emission_plot")
)

Multiple selection - Checkboxes

Modify the UI function

# Here goes everything that should appear in the app and the user can interact with
app_ui = ui.page_fluid(
    # Add title
    "Emissions Aotearoa",
    # Add selection of industry to display
    #ui.input_radio_buttons("industry_check", "Select industry", sectors), # Use this for radio buttons
    ui.input_checkbox_group("industry_check", "industry", sectors),        # Use this for checkbox
    ui.output_plot("emission_plot"),
    # Add table
    ui.output_table("emissions_table")
)

Add this to the server function

def _():
    # Always re-set the table to the full dataset, otherwise we break the dataframe
    df_table.set(df)
    #new_val = df_table()[df_table()["anzsic_descriptor"] == input.industry_check()]     # Use this for radio buttons
    new_val = df_table()[df_table()["anzsic_descriptor"].isin(input.industry_check())]   # Use this for checkbox
    df_table.set(new_val)

Let’s play with it!

Display the mean emission for each select category
Give the user the ability to set the colours for the plots
Load the extended dataset. Let the user decide which emissions to show
Whatever you’d like to do