Data Visualization

Introduction

Data visualization is a powerful tool for analyzing and presenting your research findings. Cnidarity's visualization system allows you to create interactive charts directly from your model data, helping you identify patterns, trends, and relationships in your research.

The visualization system supports multiple chart types optimized for different kinds of analysis, from showing distributions and proportions to exploring correlations between variables. All charts are generated in real-time from your current data and can be customized to highlight specific insights.

Accessing Visualizations

To access the visualization features in your project:

1. Navigate to your project: Go to the project dashboard where you want to create visualizations.
2. Open the Visualizations section: Click on "Visualizations" in the project navigation menu.
3. View existing charts: Any previously created visualizations will be displayed here.
4. Create new visualizations: Click the "Create Visualization" button to start building a new chart.

Chart Types

Cnidarity supports several chart types, each optimized for different types of data analysis:

Creating Charts

Follow these steps to create a new visualization:

1. Start the creation process:
   • Click "Create Visualization" from the Visualizations page
   • Enter a descriptive title for your chart
2. Select your data model:
   • Choose the model containing the data you want to visualize
   • Only models with records will be available for selection
3. Choose chart type:
   • Select the chart type that best fits your analysis needs
   • Available attribute options will update based on your choice
4. Configure axes:
   • For most charts: Select categorical data for X-axis, numerical data for Y-axis
   • For scatter plots: Select numerical data for both X and Y axes
   • Available attributes are filtered based on chart type requirements
5. Preview and save:
   • Use the preview to see your chart with live data
   • Adjust settings if needed
   • Click "Create Visualization" to save

Working with Scatter Plots

Scatter plots are particularly important in scientific research for analyzing relationships between continuous variables. They have special requirements and offer unique insights.

Data Requirements

• Both axes must be numerical: Unlike other chart types, scatter plots require numerical attributes for both X and Y axes
• Continuous data works best: Measurements, calculated values, and continuous scales provide the most meaningful plots
• Sufficient data points: At least 10-15 data points are recommended for meaningful pattern recognition

What Scatter Plots Reveal

• Correlations: Positive, negative, or no correlation between variables
• Outliers: Data points that don't fit the general pattern
• Clusters: Groups of similar data points
• Data distribution: How your data is spread across the value ranges

Scientific Applications

Chart Configuration

Each chart can be customized to highlight specific aspects of your data:

Title and Labels

• Chart title: Provide a clear, descriptive title that explains what the chart shows
• Axis labels: Labels are automatically generated from your attribute names
• Units: If your attributes include unit information, it will be displayed appropriately

Interactive Features

• Hover information: Hover over data points to see exact values
• Legend display: Legends are shown when helpful, hidden when they would clutter the view
• Responsive design: Charts automatically adjust to different screen sizes

Editing Visualizations

To modify an existing visualization:

1. Navigate to the Visualizations page
2. Click the "Edit" button on the visualization you want to modify
3. Update the title, chart type, model, or axis selections as needed
4. Use the preview to verify your changes
5. Click "Update Visualization" to save your changes

Requirements & Limitations

Understanding the requirements and limitations helps you make the most of the visualization system:

Subscription Requirements

Data Requirements

• Records with data: Models must contain records with actual values to generate charts
• Appropriate attribute types: Charts require numerical attributes for Y-axis (and both axes for scatter plots)
• Data variety: More diverse data creates more interesting and useful visualizations

Visualization Limits

Each subscription tier has a limit on how many visualizations each user can create per project:

• Per-user limits: Each team member can create their own visualizations up to the tier limit
• Project-specific: Limits apply separately to each project you're working on
• Real-time tracking: The system shows your current usage (e.g., "3 of 5 used") on visualization pages
• Upgrade prompts: When you reach your limit, workspace owners will see upgrade options

Current Limitations

• Single model per chart: Each visualization uses data from one model only
• Real-time data only: Charts reflect current data; historical snapshots are not available
• Attribute-based grouping: Data grouping is based on attribute values, not custom categories

Best Practices

Follow these recommendations to create effective and meaningful visualizations:

Choosing Chart Types

• Match chart to data: Use bar charts for categories, line charts for trends, scatter plots for correlations
• Consider your audience: Choose chart types that will be familiar to your intended viewers
• Avoid pie charts for many categories: Limit pie/doughnut charts to 5-7 categories maximum

Data Preparation

• Clean your data: Ensure records have consistent, accurate values before visualizing
• Use meaningful attribute names: Clear attribute names become axis labels automatically
• Include units in attribute descriptions: This helps create more informative visualizations

Creating Effective Titles

• Be descriptive: "Temperature vs. Growth Rate" is better than "Chart 1"
• Include context: Add timeframe, sample size, or experimental conditions when relevant
• Keep it concise: Aim for titles that fit comfortably above the chart

Analysis and Interpretation

• Look for patterns: Use visualizations to identify trends you might miss in raw data
• Question outliers: Investigate unusual data points - they might reveal important insights
• Update regularly: Refresh visualizations as you add new data to see evolving patterns