Figure Friday 2024 - week 42

Dash Python
1

Did you know that snakes smell by flicking their tongues in the air to pick up chemical particles?

Week 42 of Figure Friday will pay homage to our dear reptiles by learning more about them and analyzing their unique features, using the snake data set.

Thank you to Sascha Steinhoff for offering the data on the total body length (TBL), fang position, pupil shape, mode of reproduction, and toxicity of thousands of snake species.

Things to consider:

  • can you improve the sample figure built?
  • would a different figure tell the data story better?
  • can you create a Dash app instead?

Sample figure:

image
image1508×838 73.8 KB

Code for sample figure: 
import plotly.express as px
import pandas as pd

df = pd.read_csv("https://raw.githubusercontent.com/plotly/Figure-Friday/refs/heads/main/2024/week-42/merged_snake_data.csv")
df.dropna(subset=['Dentition','Pupil','Common Name'], inplace=True)

fig = px.sunburst(df, path=['Family', 'Dentition', 'Pupil','Common Name'], values='TBL cm', color='TBL cm')
fig.show()

Participation Instructions:

  • Create - use the weekly data set to build your own Plotly visualization or Dash app. Or, enhance the sample figure provided in this post, using Plotly or Dash.
  • Submit - post your creation to LinkedIn or Twitter with the hashtags #FigureFriday and #plotly by midnight Thursday, your time zone. Please also submit your visualization as a new post in this thread.
  • Celebrate - join the Figure Friday sessions to showcase your creation and receive feedback from the community.

:point_right: If you prefer to collaborate with others on Discord, join the Plotly Discord channel .

Thank you to SnakeDB for the data.

3

I chose to focus my week 42 contribution on the snake that shares its name with our favorite programming language, the Python.

The bar chart on the left is a pareto showing this world’s longest pythons. The scatter plot on the right shows a loose relation between length and weight. A polynomial curve overlay is added to show the trend. The variable degree, set to 5, is fun to experiment and see its effect on the curve. Hover info shows the common name, and gives height in cm and feet/inches, and weight in grams and pounds.

image
image793×296 58.8 KB

Here is the code:

import plotly.express as px
import plotly.graph_objects as go
from plotly.subplots import make_subplots

import polars as pl
import polars.selectors as cs

import numpy as np
from scipy.optimize import curve_fit

degree = 5  # used for curve fitting

#
#   MAKE DATAFRAMES
#
df_python = (  # df python is used for scatter plot
    pl.scan_csv('merged_snake_data.csv')
    # when Common Name is missing, use value from Binomial column 
    .with_columns(
        pl.when(pl.col('Common Name').is_null())
          .then('Binomial')
          .otherwise('Common Name')
          .alias('Common Name')
    )
    .with_columns(pl.col('Common Name').str.to_titlecase())

    .filter(pl.col('Family') == 'Pythonidae')
    .select(pl.col('TBL cm','Weight gr','Common Name'))
    .drop_nulls(subset=['TBL cm','Weight gr'])
    # add columns for hover to show length in feet and inches
    .with_columns(TOTAL_INCHES = (pl.col('TBL cm')*0.39370079))
    .with_columns(FEET = (pl.col('TOTAL_INCHES')/12).cast(pl.UInt8))
    .with_columns(INCHES = (pl.col('TOTAL_INCHES') - (12.0*pl.col('FEET'))))
    .with_columns(POUNDS = (pl.col('Weight gr') / 453.59237)) #.cast(pl.UInt8))
    .sort('TBL cm', descending=False)
    .collect()
)

# df_python_longest used for pareto chart showing world's longest pythons
df_python_longest = (
    df_python.sort('TBL cm',descending=True)
    .head(15)
    .select(pl.col('Common Name', 'TBL cm'))
    .sort('TBL cm', descending=False)
)

#
#   SETUP SIDE BY SIDE PLOTS, PARETO ON LEFT, SCATTER ON RIGHT
#
left_title = "Length of the world's longest pythons"
right_title = 'Python BMI Data: Weight vs Height<br>'
right_title += f'<sup>Best-fit polynomial degree of {degree}</sup>'
fig = make_subplots(
    rows=1, cols=2, 
    subplot_titles=(left_title, right_title),
    specs=[
            [
            {'type': 'scatter'}, {'type': 'bar'}
            ]
        ]
    )

#
#   PARETO CHART OF PYTHON LENGTHS IN LEFT PLOT
#
fig.add_trace(
    go.Bar(
        x=df_python_longest['TBL cm'],  
        y=df_python_longest['Common Name'],
        orientation='h',
        hovertemplate = None,
        hoverinfo = 'skip',
    ),
    row=1, col=1,
)
fig.update_xaxes(title_text='Length (cm)', row=1, col=1)

#
#   SCATTER PLOT WITH POLYNOMIAL BEST FIT ON THE RIGHT
#

#   Create best fit data to overlay on scatter
x = df_python['TBL cm']
y = df_python['Weight gr']
# least squares polynomial fit, 
coefs = np.polyfit(x, y, degree)
best_fit_x = np.linspace(min(x), max(x), num=np.size(x))
best_fit_y = np.polyval(coefs, best_fit_x)

fig.add_trace(
    go.Scatter(
        x=df_python['TBL cm'],
        y=df_python['Weight gr'],
        mode='markers',
        marker=dict(color='green'),
        customdata=df_python[['Common Name', 'TBL cm', 'FEET', 'INCHES', 'Weight gr', 'POUNDS']],
        hovertemplate=(
            '<b>%{customdata[0]}</b><br>' +
            'Length: %{customdata[1]} cm ' +
            '(%{customdata[2]} ft, ' +
            '%{customdata[3]:.0f} in)<br>' +
            'Weight: %{customdata[4]:,} gr ' +
            '(%{customdata[5]:.1f} Pounds)<br>' +
            '<extra></extra>'
        ),
    ),
    row=1, col=2
)
fig.update_xaxes(title_text='Length (cm)', row=1, col=2)
fig.update_yaxes(title_text='Weight (grams)', row=1, col=2)

fig.add_trace(
    go.Scatter(
        x=best_fit_x, 
        y=best_fit_y,
        hovertemplate = None,
        hoverinfo = 'skip',
        mode='lines',
        line=dict(color='gray', width=3)
    ),
    row=1, col=2
)

fig.update_layout(
    template='simple_white',
    showlegend=False,
    title=f'<b>Snakes on a Pane</b>',
    title_font={"size": 36}
)

fig.show()
fig.write_html('snakes_on_a_pane.html')
2 Likes