Data Visualization Studio

Create publication-ready, interactive data visualizations with comprehensive charting capabilities and professional styling.

When to Use This Skill

Use this skill when the user requests:

• Creating charts or graphs from data
• Building interactive dashboards
• Visualizing CSV, Excel, or JSON data
• Comparing data trends or patterns
• Exploring correlations or distributions
• Creating business intelligence visualizations
• Generating reports with visual analytics

Quick Start Workflow

1. Understand requirements - Determine chart type, data source, and styling preferences
2. Prepare data - Use preprocess_data.py for cleaning and transformation
3. Create visualization - Use create_chart.py or write custom Plotly code
4. Apply styling - Reference styling.md for professional themes
5. Export - Generate HTML for interactivity or PNG for static images

Scripts

create_chart.py

Generate interactive charts from DataFrames or CSV files with minimal code.

Supported chart types:

• line - Time series and trends
• bar - Categorical comparisons
• scatter - Relationships and correlations
• pie - Composition (use sparingly)
• histogram - Distributions
• box - Statistical distributions
• heatmap - 2D patterns and correlations
• area - Cumulative values over time

Command line usage:

python scripts/create_chart.py data.csv line date revenue output.html

Python usage:

import pandas as pd
from scripts.create_chart import create_chart, save_chart

# Load data
df = pd.read_csv('data.csv')

# Create chart
fig = create_chart(
    data=df,
    chart_type='line',
    x='date',
    y='revenue',
    color='category',  # Optional grouping
    title='Revenue Over Time',
    theme='plotly'  # or 'plotly_dark', 'seaborn', 'simple_white'
)

# Save chart
save_chart(fig, 'chart.html', format='html')

Advanced options:

• Add size parameter for scatter plots to scale point sizes
• Pass additional kwargs for chart-specific customization
• Use color for grouping by category
• Apply different themes for various contexts

preprocess_data.py

Handle data cleaning, transformation, and aggregation before visualization.

Key functions:

from scripts.preprocess_data import *

# Load data from various formats
df = load_data('data.csv')  # Also supports .xlsx, .json, .parquet

# Clean data
df = clean_data(df, drop_na=False, fill_value=0)

# Transform dates
df = transform_dates(df, date_columns=['created_at', 'updated_at'])

# Aggregate data
df_agg = aggregate_data(
    df,
    group_by=['category', 'region'],
    agg_dict={'revenue': 'sum', 'orders': 'count'},
    sort_by='revenue',
    ascending=False
)

# Create pivot table for heatmaps
pivot = create_pivot_table(df, index='month', columns='category', values='sales')

# Calculate moving averages for time series
df = calculate_moving_average(df, column='daily_sales', window=7)

# Bin numerical data
df = bin_numerical(df, column='age', bins=5, labels=['18-25', '26-35', '36-45', '46-55', '56+'])

Command line usage:

python scripts/preprocess_data.py input.csv output.csv

Reference Guides

Chart Types (references/chart_types.md)

Comprehensive guide on selecting appropriate chart types for different data scenarios. Read this when:

• User's chart type choice is unclear
• Need guidance on which visualization best suits the data
• Comparing options for a specific analysis need
• Understanding when to use specialized charts

Key sections:

• Chart selection guide with use cases
• Best practices for each chart type
• Common chart combinations for dashboards
• Drill-down patterns for interactive analysis

Styling (references/styling.md)

Professional styling and theming guidelines. Read this when:

• Creating publication-ready visualizations
• Building branded dashboards
• Ensuring accessibility and readability
• Exporting high-quality images
• Applying consistent design across multiple charts

Key sections:

• Color palettes (categorical, sequential, diverging)
• Typography recommendations
• Layout templates (minimal, dark mode, professional)
• Customization patterns (hovers, grids, annotations)
• Accessibility guidelines
• Export settings for different formats

Assets

dashboard_template.html

Professional dashboard template with responsive design. Use this when:

• Building multi-chart dashboards
• Creating executive summaries
• Presenting KPIs alongside visualizations
• Need a quick start for dashboard layouts

Features:

• Responsive grid layout
• KPI card sections
• Four chart placeholders
• Professional styling
• Mobile-friendly design

To use:

1. Copy template to working directory
2. Modify KPI values and labels
3. Replace chart data with actual data
4. Customize colors and branding as needed

Common Patterns

Pattern 1: Single Chart from CSV

import pandas as pd
from scripts.create_chart import create_chart, save_chart

df = pd.read_csv('sales_data.csv')
fig = create_chart(df, 'bar', x='product', y='revenue', title='Product Revenue')
save_chart(fig, 'product_revenue.html')

Pattern 2: Time Series with Moving Average

from scripts.preprocess_data import load_data, calculate_moving_average
from scripts.create_chart import create_chart, save_chart
import plotly.graph_objects as go

df = load_data('daily_metrics.csv')
df = calculate_moving_average(df, 'value', window=7)

# Create figure with both raw and smoothed data
fig = go.Figure()
fig.add_trace(go.Scatter(x=df['date'], y=df['value'], name='Daily', line=dict(color='lightgray')))
fig.add_trace(go.Scatter(x=df['date'], y=df['value_ma7'], name='7-day MA', line=dict(color='#636EFA', width=3)))

save_chart(fig, 'time_series.html')

Pattern 3: Multi-Category Dashboard

from scripts.preprocess_data import load_data, aggregate_data
from scripts.create_chart import create_chart
import plotly.graph_objects as go
from plotly.subplots import make_subplots

df = load_data('sales.csv')

# Create multi-chart dashboard
fig = make_subplots(
    rows=2, cols=2,
    subplot_titles=('Revenue by Region', 'Top Products', 'Monthly Trend', 'Category Mix')
)

# Add charts to subplots
# ... (add traces to each subplot)

fig.update_layout(height=800, showlegend=True)
fig.write_html('dashboard.html')

Pattern 4: Correlation Heatmap

from scripts.preprocess_data import load_data
import plotly.express as px

df = load_data('metrics.csv')
correlation = df.corr()

fig = px.imshow(
    correlation,
    labels=dict(color="Correlation"),
    x=correlation.columns,
    y=correlation.columns,
    color_continuous_scale='RdBu',
    aspect='auto'
)

fig.update_layout(title='Metric Correlations')
fig.write_html('correlation_heatmap.html')

Best Practices

Data Preparation

• Always clean data before visualization (handle nulls, duplicates)
• Transform dates to proper datetime format
• Aggregate large datasets to reduce noise
• Consider binning continuous variables for clarity

Chart Selection

• Use line charts for trends over time
• Use bar charts for categorical comparisons
• Use scatter plots to show relationships
• Limit pie chart categories to 5 or fewer
• Consider audience and purpose when selecting type

Styling

• Apply consistent color schemes across dashboard
• Ensure text is readable (proper font sizes)
• Use white space effectively
• Add titles, axis labels, and legends
• Include annotations for key insights
• Test colorblind accessibility

Performance

• Aggregate data before plotting large datasets
• Use appropriate chart types (avoid complex charts for simple data)
• Consider file size for web dashboards
• Optimize hover templates for clarity

Interactivity

• Enable hover information with relevant details
• Use appropriate mode bar options
• Consider adding drill-down capabilities
• Test responsiveness on different screen sizes

Troubleshooting

Issue: Chart appears empty or shows no data

• Verify data is loaded correctly (df.head())
• Check column names match parameters
• Ensure data types are appropriate (dates as datetime, numbers as numeric)

Issue: Colors look washed out or unclear

• Reference styling.md for better color palettes
• Ensure sufficient contrast between elements
• Test with colorblind simulation

Issue: Labels overlap or are cut off

• Adjust margins in layout
• Rotate x-axis labels if needed
• Consider abbreviating long labels
• Increase figure size

Issue: Chart is slow or unresponsive

• Reduce data points (aggregate or sample)
• Simplify chart type
• Disable unnecessary features (grid, annotations)

Example Queries This Skill Handles

• "Create a line chart showing sales over time"
• "Build a dashboard with revenue metrics"
• "Visualize the correlation between price and demand"
• "Show me a breakdown of customers by region"
• "Make an interactive chart from this CSV"
• "Create a professional-looking bar chart comparing products"
• "Analyze trends in this time series data"
• "Build a heatmap of weekly patterns"