Tidy Tuesday - Trash Wheel Collection Data

0. Loading Libraries

import pandas as pd
import ssl
from dfply import *
from plotnine import *
import numpy as np

from sklearn.preprocessing import StandardScaler
import seaborn as sns
import matplotlib.pyplot as plt

1. Data Loading

## adding below ssl line as it is giving ssl error locally in my machine
ssl._create_default_https_context = ssl._create_unverified_context

trashwheels = pd.read_csv('https://raw.githubusercontent.com/rfordatascience/tidytuesday/master/data/2024/2024-03-05/trashwheel.csv')

2. Plotting

2.1. Trash Weight & Volume Over Years

plot =  ( trashwheels >> 
        group_by(X.Year) >>
        summarize(Weights=np.sum(X.Weight), Volumes=np.sum(X.Volume)) >> 
        ggplot(aes(x = "Year",y="Weights",fill="Volumes")) + 
        geom_col() +
        theme_minimal()
        )

plot.save("TrashWeightVolume.png", width=8, height=6)

Findings:

• Consistent Growth: Trash weight and its corresponding volume consistently show an upward trend, indicating a year-on-year increase.
• Post-COVID Surge: Particularly noteworthy is the exponential rise in trash weight observed post the COVID-19 pandemic, highlighting a substantial increase in waste generation during this period.

2.2. Equivalent Homes powered with Trash Dump

plot = ( trashwheels >> 
 group_by(X.Year) >>
 summarize(HomesPowered=np.sum(X.HomesPowered)) >> 
 ggplot(aes(x = "Year",y="HomesPowered")) + 
 geom_line(color = "brown") +
 theme_minimal()
 )

plot.save("HomePowered.png", width=8, height=6)

Findings:

• Intuitive Observation: It is evident from the data that the number of homes powered by trash is intuitively increasing over time.

2.3. Trash by Category Trending

plot =  (trashwheels >>
        gather("variable", "value", ["PlasticBottles", "Polystyrene", "CigaretteButts", "GlassBottles", "PlasticBags", "Wrappers", "SportsBalls"]) >>
        group_by(X.Year, X.variable) >>
        summarize(Values=np.sum(X.value)) >>
        group_by(X.variable) >>
        mutate(Standardized_Values=(X.Values - X.Values.mean()) / X.Values.std()) >>
        ggplot(aes(x="Year", y="Standardized_Values")) +
        geom_line(color="brown") +
        facet_wrap('~variable') +
        theme_minimal()
        )

plot.save("TrashVolumeCat.png", width=8, height=6)

Findings:

• Increasing Usage and Disposal: The consumption and disposal of Glass Bottles, Plastic Bottles, Sport Balls, and Plastic Wrappers have witnessed a steady increase in recent years.
• Processed Food Consumption: The elevated consumption of Plastic, Wrappers, and Glass suggests a potential upward trend in the usage of processed food items, given their common packaging materials.
• Quality Impact on Dumping: The higher consumption of Sport Balls may be linked to quality factors, potentially indicating a shorter lifespan for these items than expected. This, in turn, could contribute to an increase in waste disposal.

2.4. Correlation Heat Map B/w Trash Weight & Content

selected_columns = ["Weight", "PlasticBottles", "Polystyrene", "CigaretteButts", "GlassBottles", "PlasticBags", "Wrappers", "SportsBalls"] 
selected_data = trashwheels[selected_columns]

scaler = StandardScaler()
scaled_data = scaler.fit_transform(selected_data)

scaled_df = pd.DataFrame(scaled_data, columns=selected_columns)

correlation_matrix_scaled = scaled_df.corr()

plt.figure(figsize=(6, 4))
sns.heatmap(correlation_matrix_scaled, annot=True, cmap='coolwarm', fmt=".2f", linewidths=.5)
plt.title('Correlation Heatmap (Scaled Data)')
plt.savefig('correlation_heatmap.png')

Findings:

• Limited Observations: Upon examination, I did not observe a significant association as expected. However, a noticeable correlation exists between the usage of plastic bottles and related products such as plastic bags, wrappers, and polystyrene.
• Material Association: The correlation between plastic bottles and sport balls suggests a potential association in the manufacturing process, possibly sharing similar synthetic raw materials.