Code Editor for app.py

import streamlit as st
import pandas as pd
import numpy as np
import altair as alt
import seaborn as sns

# Cache the data loading function
@st.cache_data
def load_data(url):
    try:
        return pd.read_csv(url)
    except Exception as e:
        st.error(f"Error loading data: {e}")
        return None

# URL input with default value
url = st.text_input(
    "Enter CSV URL",
    value="https://raw.githubusercontent.com/mwaskom/seaborn-data/master/penguins.csv"
)

data = load_data(url)

if data is not None:
    # Controls for data display
    col1, col2 = st.columns(2)
    
    with col1:
        n_rows = st.number_input("Number of rows to display", 
                                min_value=1, 
                                max_value=len(data), 
                                value=min(10, len(data)))
    
    with col2:
        shuffle = st.checkbox("Randomly shuffle data")
    
    # Shuffle data if requested
    if shuffle:
        data = data.sample(frac=1, random_state=42).reset_index(drop=True)
    
    # Display the dataframe
    st.subheader("Data Preview")
    st.dataframe(data.head(n_rows))
    
    # Variable selection for visualization
    st.subheader("Visualization")
    variables = list(data.select_dtypes(include=[np.number, 'category', 'object']).columns)
    
    col1, col2 = st.columns(2)
    with col1:
        var1 = st.selectbox("Select first variable", variables)
    with col2:
        var2 = st.selectbox("Select second variable", 
                           [None] + [v for v in variables if v != var1])
    
    if var2 is None:
        # Single variable visualization
        if pd.api.types.is_numeric_dtype(data[var1]):
            # Histogram for numeric variables
            chart = alt.Chart(data).mark_bar().encode(
                alt.X(f"{var1}:Q", bin=True),
                y='count()'
            ).properties(
                title=f"Distribution of {var1}"
            )
        else:
            # Bar chart for categorical variables
            chart = alt.Chart(data).mark_bar().encode(
                x='count()',
                y=alt.Y(f"{var1}:N", sort='-x')
            ).properties(
                title=f"Distribution of {var1}"
            )
    else:
        # Two variable visualization
        if (pd.api.types.is_numeric_dtype(data[var1]) and 
            pd.api.types.is_numeric_dtype(data[var2])):
            # Scatter plot for two numeric variables
            chart = alt.Chart(data).mark_circle().encode(
                x=f"{var1}:Q",
                y=f"{var2}:Q"
            ).properties(
                title=f"{var1} vs {var2}"
            )
        elif (pd.api.types.is_numeric_dtype(data[var1]) and 
              not pd.api.types.is_numeric_dtype(data[var2])):
            # Box plot for numeric + categorical
            chart = alt.Chart(data).mark_boxplot().encode(
                x=f"{var2}:N",
                y=f"{var1}:Q"
            ).properties(
                title=f"{var1} by {var2}"
            )
        elif (not pd.api.types.is_numeric_dtype(data[var1]) and 
              pd.api.types.is_numeric_dtype(data[var2])):
            # Box plot for categorical + numeric
            chart = alt.Chart(data).mark_boxplot().encode(
                x=f"{var1}:N",
                y=f"{var2}:Q"
            ).properties(
                title=f"{var2} by {var1}"
            )
        else:
            # Heatmap for two categorical variables
            chart = alt.Chart(data).mark_rect().encode(
                x=f"{var1}:N",
                y=f"{var2}:N",
                color='count()'
            ).properties(
                title=f"Frequency of {var1} vs {var2}"
            )
    
    st.altair_chart(chart, use_container_width=True)