Understanding Results

As Querri analyzes your data, it produces results in various formats. This guide helps you understand and work with these results effectively.

Types of Results

Querri can display results in three primary formats:

Table Results (QDF)

Tabular data appears in the QDF (Querri Data Frame) viewer:

• Interactive scrolling: Navigate through rows and columns
• Column headers: Click to understand data types
• Clean formatting: Numbers, dates, and text are properly formatted
• Large dataset support: Handles thousands of rows smoothly

Tables are the most common result type, used for:

• Raw data previews
• Filtered datasets
• Aggregated summaries
• Joined or transformed data

Charts and Visualizations

Visual results appear as interactive charts:

• Multiple chart types: Bar, line, scatter, area, pie, and more
• Readable legends: Understand what each element represents
• Tooltips: Hover for detailed values
• Responsive sizing: Charts adapt to your screen

Visualizations help you spot trends, patterns, and outliers quickly.

Text Insights

Some results are textual explanations:

• Summary statistics: “Average revenue: $45,234”
• Answers to questions: “There are 1,247 unique customers”
• Interpretations: “Revenue increased 23% year-over-year”
• Recommendations: “Consider filtering outliers before forecasting”

Text results provide context and explanations alongside data.

Reading the QDF Viewer

The QDF viewer is Querri’s table interface. Here’s how to use it:

Navigation

• Scroll vertically: See more rows
• Scroll horizontally: See more columns
• Pagination: For very large datasets, navigate between pages

Understanding Columns

Each column header shows:

• Column name: What the data represents
• Data type: Number, text, date, boolean, etc.
• Sort options: Click to sort ascending or descending (when available)

Cell Values

Individual cells display:

• Numbers: Formatted with appropriate decimals and separators
• Dates: In human-readable format
• Text: Truncated with ellipsis if too long (hover for full value)
• Null values: Shown as empty or “null”

Row Counts

The viewer typically shows:

• Total number of rows
• Current page or range displayed
• Any filters applied

Understanding Charts

When a step produces a visualization, here’s what to look for:

Chart Components

Title: Describes what the chart shows Axes: X-axis (horizontal) and Y-axis (vertical) with labels Legend: Explains colors, lines, or segments Data points/bars/lines: The actual visual representation

Interactive Features

Many charts support:

• Hover tooltips: Exact values when you mouse over elements
• Zoom: Click and drag to zoom into specific areas (when supported)
• Pan: Navigate around zoomed charts

Reading Common Chart Types

Line charts: Show trends over time. Look for slopes (increasing/decreasing) and inflection points.

Bar charts: Compare values across categories. Height represents magnitude.

Scatter plots: Show relationships between two variables. Patterns indicate correlation.

Pie charts: Display proportions of a whole. Size represents percentage.

Area charts: Like line charts but with filled areas, good for showing cumulative values.

Step Status Indicators

Each step in your project shows its status:

Running

Indicator: Spinner or progress animation Meaning: The step is currently executing Action: Wait for completion

Steps usually complete in seconds, though complex queries or large datasets may take longer.

Success

Indicator: Green checkmark or success icon Meaning: The step completed without errors Results: Available to view below the step

Most steps should complete successfully. Review the results to ensure they match your expectations.

Error

Indicator: Red X or error icon Meaning: The step encountered a problem and couldn’t complete Information: Error message explains what went wrong

Common causes of errors:

• Invalid column names
• Data type mismatches
• Missing data sources
• Syntax issues in complex operations

When you see an error:

1. Read the error message carefully
2. Identify what went wrong
3. Ask the AI to fix it or rephrase your question
4. Continue with your analysis

The AI can usually recover from errors and continue.

Downloading Results

You can export results for use outside Querri:

Downloading Tables

For QDF table results:

1. Look for a Download or Export button
2. Select format (typically CSV or Excel)
3. Save to your computer

Use downloaded data in:

• Spreadsheet applications
• Other analysis tools
• Reports and presentations
• Archival purposes

Downloading Charts

For visualizations:

1. Find the Download or Export option
2. Choose format (PNG, JPEG, SVG, or PDF)
3. Save the image

Use exported charts in:

• Presentations
• Reports
• Dashboards
• Publications

Downloading Full Projects

Some interfaces let you export entire project:

• All steps and their results
• Data sources used
• Complete analysis narrative

This is useful for:

• Archiving completed work
• Sharing with non-Querri users
• Regulatory compliance

Copying Data

For quick access to results:

Copy Table Data

1. Select cells in the QDF viewer (if supported)
2. Right-click and Copy, or use Ctrl+C / Cmd+C
3. Paste into spreadsheets, documents, or other tools

Copy Chart Images

1. Right-click on a chart
2. Select “Copy image” or similar option
3. Paste into documents or image editors

Copy Text Results

Simply select and copy text insights like any other text on the web.

Interpreting Results in Context

Understanding individual results is important, but so is seeing the bigger picture:

Sequential Steps

Results build on each other:

• Step 1 loads data (1,000 rows)
• Step 2 filters to 2024 (800 rows)
• Step 3 aggregates by month (12 rows)
• Step 4 creates chart (visual of 12 data points)

Watch how data transforms through the sequence.

Validating Results

Ask yourself:

• Do these numbers make sense? Check for obvious errors.
• Is this what I expected? Unexpected results might reveal insights or issues.
• Should I verify? Cross-reference with source data when critical.

Following the Thread

In a conversational analysis:

• Each result answers part of your question
• Results may raise new questions
• The conversation builds understanding incrementally

Don’t just look at the final result—the journey provides context.

Common Result Patterns

Empty Results

If a step returns no rows:

• Your filter might be too restrictive
• There might be a data type issue (e.g., date format)
• The data simply might not contain what you’re looking for

Action: Broaden your filter or check your data.

Unexpected Row Counts

If you get far more or fewer rows than expected:

• Check for duplicates (more rows)
• Verify your filter conditions (fewer rows)
• Look for join issues (often multiplies rows)

Action: Ask the AI to explain the row count or adjust your query.

Strange Values

If you see odd numbers, dates, or text:

• Data types might be misinterpreted
• Source data might have quality issues
• Calculations might need adjustment

Action: Use data cleaning steps or ask the AI to handle the issue.

Best Practices

Review Each Step

Don’t skip ahead—look at each step’s results:

• Catch errors early
• Understand the analysis flow
• Verify assumptions

Spot Check Data

For important analyses:

• Compare a few results to source data manually
• Check that aggregations make sense
• Verify calculations with a calculator

Ask Questions About Results

If something looks off:

• “Why are there only 3 rows here?”
• “What does this null value mean?”
• “Can you explain this outlier?”

The AI can help interpret results.

Save Important Results

Don’t rely solely on the project:

• Download critical tables
• Export key visualizations
• Take screenshots of important insights

This creates a backup and makes sharing easier.

Next Steps

• Learn how to create better visualizations
• Explore project management to organize your work
• Understand sharing options to collaborate on results