数据星期二

A guide to the TidyTuesday data project, demonstrating how to access data and create interactive Shiny apps in both R and Python for data exploration and visualization.

This document provides a comprehensive guide to the TidyTuesday data project, demonstrating how to access and work with the data in both R and Python. It also includes detailed instructions on how to create interactive Shiny apps for data exploration and visualization. The guide covers everything from downloading the data to building a complete Shiny app with various plots and user inputs. This is a valuable resource for anyone looking to participate in the TidyTuesday project and improve their data science skills.

Explores the TidyTuesday data project, demonstrating data access and interactive Shiny app creation in R and Python.

TidyTuesday data project

data from github

getting the data

R

#pak::pak('tidytuesdayR')

library(tidytuesdayR)
library(tidyverse)

download the data

all available data

#tt_available() 

tuesdata <- tidytuesdayR::tt_load('2025-04-01')

tuesdata

Python

import pandas as pd
import pydytuesday

pydytuesday.get_date('2025-04-15')

make a shiny

R

read data

data=tuesdata$pokemon_df
#glimpse(data)

or read directly from the url

data<- readr::read_csv('https://raw.githubusercontent.com/rfordatascience/tidytuesday/main/data/2025/2025-04-01/pokemon_df.csv')

using shinyapps assistant to create shinyapp

https://gallery.shinyapps.io/assistant

go to project folder and install quarto-ext/shinylive

quarto add quarto-ext/shinylive

Prompt:

create a shinyapp with this data from github:https://raw.githubusercontent.com/rfordatascience/tidytuesday/main/data/2025/2025-04-01/pokemon_df.csv. left side: selector type_1 number of pokemon right side: histogram of height, color by type_2 histogram of weight, color by type_2 histogram of attack, color by type_2 histogram of defense color by type_2 barplot of color_1

there is no weight_kg,height_m.please use correct name.

Shiny R in quarto

if adding shiny in quarto then adding this to yaml header

---

filters:
  - shinylive
---

library(shiny)
library(bslib)
library(tidyverse)
library(ggplot2)

# Load the Pokemon data
pokemon_data <- read.csv("https://raw.githubusercontent.com/rfordatascience/tidytuesday/main/data/2025/2025-04-01/pokemon_df.csv")

ui <- page_sidebar(
  title = "Pokemon Data Explorer",
  sidebar = sidebar(
    selectInput("type1", "Select Type 1:", 
                choices = sort(unique(pokemon_data$type_1)),
                selected = "water"),
    hr(),
    textOutput("pokemon_count")
  ),
  
  layout_columns(
    card(
      card_header("Height Distribution by Type 2"),
      plotOutput("height_hist")
    ),
    card(
      card_header("Weight Distribution by Type 2"),
      plotOutput("weight_hist")
    )
  ),
  
  layout_columns(
    card(
      card_header("Attack Distribution by Type 2"),
      plotOutput("attack_hist")
    ),
    card(
      card_header("Defense Distribution by Type 2"),
      plotOutput("defense_hist")
    )
  ),
  
  card(
    card_header("Pokemon Color Distribution"),
    plotOutput("color_barplot")
  )
)

server <- function(input, output, session) {
  
  # Filtered data based on the selected type_1
  filtered_data <- reactive({
    pokemon_data %>%
      filter(type_1 == input$type1)
  })
  
  # Display number of Pokemon
  output$pokemon_count <- renderText({
    count <- nrow(filtered_data())
    paste("Number of Pokemon with Type 1 '", input$type1, "': ", count)
  })
  
  # Height histogram colored by type_2
  output$height_hist <- renderPlot({
    ggplot(filtered_data(), aes(x = height, fill = type_2)) +
      geom_histogram(alpha = 0.7, bins = 20, position = "identity") +
      scale_fill_viridis_d() +
      theme_minimal() +
      labs(x = "Height", y = "Count", fill = "Type 2")
  })
  
  # Weight histogram colored by type_2
  output$weight_hist <- renderPlot({
    ggplot(filtered_data(), aes(x = weight, fill = type_2)) +
      geom_histogram(alpha = 0.7, bins = 20, position = "identity") +
      scale_fill_viridis_d() +
      theme_minimal() +
      labs(x = "Weight", y = "Count", fill = "Type 2")
  })
  
  # Attack histogram colored by type_2
  output$attack_hist <- renderPlot({
    ggplot(filtered_data(), aes(x = attack, fill = type_2)) +
      geom_histogram(alpha = 0.7, bins = 20, position = "identity") +
      scale_fill_viridis_d() +
      theme_minimal() +
      labs(x = "Attack", y = "Count", fill = "Type 2")
  })
  
  # Defense histogram colored by type_2
  output$defense_hist <- renderPlot({
    ggplot(filtered_data(), aes(x = defense, fill = type_2)) +
      geom_histogram(alpha = 0.7, bins = 20, position = "identity") +
      scale_fill_viridis_d() +
      theme_minimal() +
      labs(x = "Defense", y = "Count", fill = "Type 2")
  })
  
  # Barplot of color_1
  output$color_barplot <- renderPlot({
    color_counts <- filtered_data() %>%
      count(color_1) %>%
      arrange(desc(n))
    
    ggplot(color_counts, aes(x = reorder(color_1, n), y = n, fill = color_1)) +
      geom_col() +
      coord_flip() +
      scale_fill_brewer(palette = "Set3") +
      theme_minimal() +
      labs(x = "Color", y = "Count", fill = "Color") +
      theme(legend.position = "none")
  })
}

shinyApp(ui, server)

#| '!! shinylive warning !!': |
#|   shinylive does not work in self-contained HTML documents.
#|   Please set `embed-resources: false` in your metadata.
#| standalone: true
#| viewerHeight: 800
library(shiny)
library(bslib)
library(tidyverse)
library(ggplot2)

# Load the Pokemon data
pokemon_data <- read.csv("https://raw.githubusercontent.com/rfordatascience/tidytuesday/main/data/2025/2025-04-01/pokemon_df.csv")

ui <- page_sidebar(
  title = "Pokemon Data Explorer",
  sidebar = sidebar(
    selectInput("type1", "Select Type 1:", 
                choices = sort(unique(pokemon_data$type_1)),
                selected = "water"),
    hr(),
    textOutput("pokemon_count")
  ),
  
  layout_columns(
    card(
      card_header("Height Distribution by Type 2"),
      plotOutput("height_hist")
    ),
    card(
      card_header("Weight Distribution by Type 2"),
      plotOutput("weight_hist")
    )
  ),
  
  layout_columns(
    card(
      card_header("Attack Distribution by Type 2"),
      plotOutput("attack_hist")
    ),
    card(
      card_header("Defense Distribution by Type 2"),
      plotOutput("defense_hist")
    )
  ),
  
  card(
    card_header("Pokemon Color Distribution"),
    plotOutput("color_barplot")
  )
)

server <- function(input, output, session) {
  
  # Filtered data based on the selected type_1
  filtered_data <- reactive({
    pokemon_data %>%
      filter(type_1 == input$type1)
  })
  
  # Display number of Pokemon
  output$pokemon_count <- renderText({
    count <- nrow(filtered_data())
    paste("Number of Pokemon with Type 1 '", input$type1, "': ", count)
  })
  
  # Height histogram colored by type_2
  output$height_hist <- renderPlot({
    ggplot(filtered_data(), aes(x = height, fill = type_2)) +
      geom_histogram(alpha = 0.7, bins = 20, position = "identity") +
      scale_fill_viridis_d() +
      theme_minimal() +
      labs(x = "Height", y = "Count", fill = "Type 2")
  })
  
  # Weight histogram colored by type_2
  output$weight_hist <- renderPlot({
    ggplot(filtered_data(), aes(x = weight, fill = type_2)) +
      geom_histogram(alpha = 0.7, bins = 20, position = "identity") +
      scale_fill_viridis_d() +
      theme_minimal() +
      labs(x = "Weight", y = "Count", fill = "Type 2")
  })
  
  # Attack histogram colored by type_2
  output$attack_hist <- renderPlot({
    ggplot(filtered_data(), aes(x = attack, fill = type_2)) +
      geom_histogram(alpha = 0.7, bins = 20, position = "identity") +
      scale_fill_viridis_d() +
      theme_minimal() +
      labs(x = "Attack", y = "Count", fill = "Type 2")
  })
  
  # Defense histogram colored by type_2
  output$defense_hist <- renderPlot({
    ggplot(filtered_data(), aes(x = defense, fill = type_2)) +
      geom_histogram(alpha = 0.7, bins = 20, position = "identity") +
      scale_fill_viridis_d() +
      theme_minimal() +
      labs(x = "Defense", y = "Count", fill = "Type 2")
  })
  
  # Barplot of color_1
  output$color_barplot <- renderPlot({
    color_counts <- filtered_data() %>%
      count(color_1) %>%
      arrange(desc(n))
    
    ggplot(color_counts, aes(x = reorder(color_1, n), y = n, fill = color_1)) +
      geom_col() +
      coord_flip() +
      scale_fill_brewer(palette = "Set3") +
      theme_minimal() +
      labs(x = "Color", y = "Count", fill = "Color") +
      theme(legend.position = "none")
  })
}

shinyApp(ui, server)

Python

install shiny in python

Need to down grade shinylive Python version to 0.7.1 in order to match shinylive R version

import os
os.system("pip install 'shinylive==0.7.1'")

read data

penguins = pd.read_csv('penguins.csv')
penguins_raw = pd.read_csv('penguins_raw.csv')

# Option 2: Read directly from GitHub and assign to an object
#penguins = pd.read_csv('https://raw.githubusercontent.com/rfordatascience/tidytuesday/main/data/2025/2025-04-15/penguins.csv')
#penguins_raw = pd.read_csv('https://raw.githubusercontent.com/rfordatascience/tidytuesday/main/data/2025/2025-04-15/penguins_raw.csv')

using shinyapps assistant to create shinyapp

https://gallery.shinyapps.io/assistant

install shinylive

pip install shinylive

Prompt:

create a shinyapp with this data from github:https://raw.githubusercontent.com/rfordatascience/tidytuesday/main/data/2025/2025-04-15/penguins_raw.csv.

species choices is not correct,please use correct choices.

Shiny Python in quarto

if adding shiny in quarto then adding this to yaml header

---

filters:
  - shinylive
---

from shiny import App, reactive, render, ui
import pandas as pd
import matplotlib.pyplot as plt
import seaborn as sns
import urllib3

app_ui = ui.page_sidebar(
    ui.sidebar(
        ui.h3("Penguin Data Explorer"),
        ui.input_select(
            "species",
            "Select Penguin Species",
            choices=["All Species", "Adelie Penguin (Pygoscelis adeliae)", 
                     "Gentoo penguin (Pygoscelis papua)", 
                     "Chinstrap penguin (Pygoscelis antarctica)"]
        ),
        ui.input_select(
            "plot_type",
            "Select Plot Type",
            choices=[
                "Body Mass vs Flipper Length",
                "Culmen Length vs Depth",
                "Histogram of Body Mass"
            ]
        ),
        ui.input_checkbox_group(
            "islands",
            "Select Islands",
            choices=["Torgersen", "Biscoe", "Dream"],
            selected=["Torgersen", "Biscoe", "Dream"]
        ),
        ui.hr(),
        ui.p("Data from Palmer Penguins dataset via TidyTuesday."),
    ),
    ui.card(
        ui.card_header("Penguin Data Visualization"),
        ui.output_plot("penguin_plot")
    ),
    ui.card(
        ui.card_header("Data Summary"),
        ui.output_table("summary_table")
    )
)

def server(input, output, session):
    # Load data
    @reactive.calc
    def load_data():
        http = urllib3.PoolManager()
        url = "https://raw.githubusercontent.com/rfordatascience/tidytuesday/main/data/2025/2025-04-15/penguins_raw.csv"
        
        try:
            response = http.request('GET', url)
            if response.status != 200:
                # If file not found, use fallback URL
                fallback_url = "https://raw.githubusercontent.com/allisonhorst/palmerpenguins/main/inst/extdata/penguins_raw.csv"
                response = http.request('GET', fallback_url)
                
            data = pd.read_csv(pd.io.common.StringIO(response.data.decode('utf-8')))
            return data
        except Exception as e:
            print(f"Error loading data: {e}")
            # Return empty dataframe if there are issues
            return pd.DataFrame()

    @reactive.calc
    def filtered_data():
        data = load_data()
        if data.empty:
            return pd.DataFrame()
        
        # Clean column names
        data.columns = [col.strip() for col in data.columns]
        
        # Filter data based on inputs
        filtered = data.copy()
        
        # Filter by species
        if input.species() != "All Species":
            filtered = filtered[filtered['Species'] == input.species()]
        
        # Filter by islands
        filtered = filtered[filtered['Island'].isin(input.islands())]
        
        return filtered

    @output
    @render.plot
    def penguin_plot():
        data = filtered_data()
        if data.empty:
            fig, ax = plt.subplots()
            ax.text(0.5, 0.5, "No data available or error loading data", 
                    ha='center', va='center')
            ax.set_xlim(0, 1)
            ax.set_ylim(0, 1)
            ax.axis('off')
            return fig
        
        fig, ax = plt.subplots(figsize=(10, 6))
        
        plot_type = input.plot_type()
        
        if plot_type == "Body Mass vs Flipper Length":
            sns.scatterplot(
                data=data, 
                x='Flipper Length (mm)', 
                y='Body Mass (g)',
                hue='Species',
                style='Sex',
                ax=ax
            )
            ax.set_title("Body Mass vs Flipper Length")
            
        elif plot_type == "Culmen Length vs Depth":
            sns.scatterplot(
                data=data, 
                x='Culmen Length (mm)', 
                y='Culmen Depth (mm)',
                hue='Species',
                style='Sex',
                ax=ax
            )
            ax.set_title("Culmen Length vs Depth")
            
        elif plot_type == "Histogram of Body Mass":
            sns.histplot(
                data=data,
                x='Body Mass (g)',
                hue='Species',
                kde=True,
                ax=ax
            )
            ax.set_title("Distribution of Body Mass")
            
        plt.tight_layout()
        return fig

    @output
    @render.table
    def summary_table():
        data = filtered_data()
        if data.empty:
            return pd.DataFrame({'Message': ['No data available or error loading data']})
        
        # Create a summary table with counts by species and island
        summary = data.groupby(['Species', 'Island', 'Sex']).size().reset_index(name='Count')
        
        # Add some descriptive statistics
        stats = data.groupby(['Species']).agg({
            'Body Mass (g)': ['mean', 'std'],
            'Flipper Length (mm)': ['mean', 'std'],
            'Culmen Length (mm)': ['mean', 'std'],
            'Culmen Depth (mm)': ['mean', 'std']
        }).round(2)
        
        stats.columns = ['_'.join(col).strip() for col in stats.columns.values]
        stats = stats.reset_index()
        
        return summary

app = App(app_ui, server)

#| '!! shinylive warning !!': |
#|   shinylive does not work in self-contained HTML documents.
#|   Please set `embed-resources: false` in your metadata.
#| standalone: true
#| viewerHeight: 800
from shiny import App, reactive, render, ui
import pandas as pd
import matplotlib.pyplot as plt
import seaborn as sns
import urllib3

app_ui = ui.page_sidebar(
    ui.sidebar(
        ui.h3("Penguin Data Explorer"),
        ui.input_select(
            "species",
            "Select Penguin Species",
            choices=["All Species", "Adelie Penguin (Pygoscelis adeliae)", 
                     "Gentoo penguin (Pygoscelis papua)", 
                     "Chinstrap penguin (Pygoscelis antarctica)"]
        ),
        ui.input_select(
            "plot_type",
            "Select Plot Type",
            choices=[
                "Body Mass vs Flipper Length",
                "Culmen Length vs Depth",
                "Histogram of Body Mass"
            ]
        ),
        ui.input_checkbox_group(
            "islands",
            "Select Islands",
            choices=["Torgersen", "Biscoe", "Dream"],
            selected=["Torgersen", "Biscoe", "Dream"]
        ),
        ui.hr(),
        ui.p("Data from Palmer Penguins dataset via TidyTuesday."),
    ),
    ui.card(
        ui.card_header("Penguin Data Visualization"),
        ui.output_plot("penguin_plot")
    ),
    ui.card(
        ui.card_header("Data Summary"),
        ui.output_table("summary_table")
    )
)

def server(input, output, session):
    # Load data
    @reactive.calc
    def load_data():
        http = urllib3.PoolManager()
        url = "https://raw.githubusercontent.com/rfordatascience/tidytuesday/main/data/2025/2025-04-15/penguins_raw.csv"
        
        try:
            response = http.request('GET', url)
            if response.status != 200:
                # If file not found, use fallback URL
                fallback_url = "https://raw.githubusercontent.com/allisonhorst/palmerpenguins/main/inst/extdata/penguins_raw.csv"
                response = http.request('GET', fallback_url)
                
            data = pd.read_csv(pd.io.common.StringIO(response.data.decode('utf-8')))
            return data
        except Exception as e:
            print(f"Error loading data: {e}")
            # Return empty dataframe if there are issues
            return pd.DataFrame()

    @reactive.calc
    def filtered_data():
        data = load_data()
        if data.empty:
            return pd.DataFrame()
        
        # Clean column names
        data.columns = [col.strip() for col in data.columns]
        
        # Filter data based on inputs
        filtered = data.copy()
        
        # Filter by species
        if input.species() != "All Species":
            filtered = filtered[filtered['Species'] == input.species()]
        
        # Filter by islands
        filtered = filtered[filtered['Island'].isin(input.islands())]
        
        return filtered

    @output
    @render.plot
    def penguin_plot():
        data = filtered_data()
        if data.empty:
            fig, ax = plt.subplots()
            ax.text(0.5, 0.5, "No data available or error loading data", 
                    ha='center', va='center')
            ax.set_xlim(0, 1)
            ax.set_ylim(0, 1)
            ax.axis('off')
            return fig
        
        fig, ax = plt.subplots(figsize=(10, 6))
        
        plot_type = input.plot_type()
        
        if plot_type == "Body Mass vs Flipper Length":
            sns.scatterplot(
                data=data, 
                x='Flipper Length (mm)', 
                y='Body Mass (g)',
                hue='Species',
                style='Sex',
                ax=ax
            )
            ax.set_title("Body Mass vs Flipper Length")
            
        elif plot_type == "Culmen Length vs Depth":
            sns.scatterplot(
                data=data, 
                x='Culmen Length (mm)', 
                y='Culmen Depth (mm)',
                hue='Species',
                style='Sex',
                ax=ax
            )
            ax.set_title("Culmen Length vs Depth")
            
        elif plot_type == "Histogram of Body Mass":
            sns.histplot(
                data=data,
                x='Body Mass (g)',
                hue='Species',
                kde=True,
                ax=ax
            )
            ax.set_title("Distribution of Body Mass")
            
        plt.tight_layout()
        return fig

    @output
    @render.table
    def summary_table():
        data = filtered_data()
        if data.empty:
            return pd.DataFrame({'Message': ['No data available or error loading data']})
        
        # Create a summary table with counts by species and island
        summary = data.groupby(['Species', 'Island', 'Sex']).size().reset_index(name='Count')
        
        # Add some descriptive statistics
        stats = data.groupby(['Species']).agg({
            'Body Mass (g)': ['mean', 'std'],
            'Flipper Length (mm)': ['mean', 'std'],
            'Culmen Length (mm)': ['mean', 'std'],
            'Culmen Depth (mm)': ['mean', 'std']
        }).round(2)
        
        stats.columns = ['_'.join(col).strip() for col in stats.columns.values]
        stats = stats.reset_index()
        
        return summary

app = App(app_ui, server)

Reference

https://github.com/rfordatascience/tidytuesday

https://github.com/posit-dev/python-tidytuesday