# ---------------------------------------------------------------------------- #

# HELPFUL SOURCES

# ---------------------------------------------------------------------------- #

# http://seaborn.pydata.org/index.html

# https://discuss.streamlit.io/t/rendering-matplotlib-axessubplots-in-streamlit/5662/5

# https://stackoverflow.com/questions/45569592/saving-plot-from-seaborn?rq=1

# https://stackoverflow.com/questions/36018681/stop-seaborn-plotting-multiple-figures-on-top-of-one-another

# https://www.datacamp.com/community/blog/seaborn-cheat-sheet-python

# https://www.kdnuggets.com/2019/04/data-visualization-python-matplotlib-seaborn.html

# https://seaborn.pydata.org/tutorial/axis_grids.html (axis grids)

# ---------------------------------------------------------------------------- #

# IMPORT PACKAGES

# ---------------------------------------------------------------------------- #

import streamlit as st

import seaborn as sns

# ---------------------------------------------------------------------------- #

st.title('Integrating Seaborn and Streamlit')

st.subheader('Rebecca Weng | Dec. 2020')

fig_type = st.selectbox('Plot Type', ('Histogram', 'Lineplot',

'Scatterplot', 'Boxplot'))

# ---------------------------------------------------------------------------- #

# Adapted from "Stacked histogram on a log scale" from seaborn.pydata.org

if fig_type == 'Histogram':

diamonds = sns.load_dataset("diamonds")

# Workaround for displaying categorical variables

st.write(diamonds.astype('object'))

sns.set_theme(style="darkgrid")

# Slider to determine the number of bins

bin_num = st.slider("Number of Bins #1", min_value = 5, max_value = 100,

value = 30, step = 5)

plt = sns.histplot(diamonds[diamonds['cut'] == 'Ideal'], x="price",

edgecolor="0.3", linewidth=.5,

log_scale=True, bins = bin_num) # AxesSubplot Object

plt.set_title("Histogram of Price of Ideal Cut Diamonds")

# Necessary line of code to get underlying plot's object

fig = plt.get_figure() # Figure object

st.pyplot(fig) # produces graph, same as st.write(fig)

# Clears figure so you can plot the next without retaining first plot

fig.clf()

# Slider to determine the number of bins

bin_num2 = st.slider("Number of Bins #2", min_value = 5, max_value = 100,

value = 30, step = 5)

plt2 = sns.histplot(diamonds, x="price", hue="cut", multiple="stack",

palette="light:m_r", edgecolor=".3", linewidth=.5,

log_scale=True, bins = bin_num2) # AxesSubplot Object

plt2.set_title("Stacked Histogram of Diamond Price by Cut")

fig2 = plt2.get_figure()

st.pyplot(fig2)

# ---------------------------------------------------------------------------- #

# Adapted from "Timseries plot with error bands" from seaborn.pydata.org

elif fig_type == 'Lineplot':

sns.set_theme(style="darkgrid")

# Load an example dataset with long-form data

fmri = sns.load_dataset("fmri")

# Plot the responses for different events and regions

plt3 = sns.lineplot(x="timepoint", y="signal",

hue="region", style="event",

data=fmri)

fig3 = plt3.get_figure()

st.pyplot(fig3)

# ---------------------------------------------------------------------------- #

# Adapted from "Scatterplot with multiple semantics" from seaborn.pydata.org

elif fig_type == 'Scatterplot':

# Load the example diamonds dataset

diamonds = sns.load_dataset("diamonds")

# Draw a scatterplot while assigning point colors and sizes to different

# variables in the dataset

clarity_ranking = ["I1", "SI2", "SI1", "VS2", "VS1", "VVS2", "VVS1", "IF"]

plt4 = sns.scatterplot(x="carat", y="price",

hue="clarity", hue_order=clarity_ranking,

palette="ch:r=-.2,d=.3_r", size="depth",

sizes=(1, 8), linewidth=0,

data=diamonds)

plt4.legend(fontsize = 'small')

fig4 = plt4.get_figure()

st.pyplot(fig4)

# ---------------------------------------------------------------------------- #

# Adapted from "Scatterplot with multiple semantics" from seaborn.pydata.org

elif fig_type == 'Boxplot':

# Load the example tips dataset

tips = sns.load_dataset("tips")

# Draw a nested boxplot to show bills by day and time

plt5 = sns.boxplot(x="day", y="total_bill",

hue="smoker", data=tips)

plt5.legend(fontsize = 'small')

plt5.set_xlabel("Total Bill")

plt5.set_ylabel("Day of the Week")

fig5 = plt5.get_figure()

st.pyplot(fig5)