*Create insight

Create insights to analyze and visualize your data in the Polyteia Platform.

Step-by-step guide

1

Open the insight editor

From a dataset

1. Navigate to the desired dataset through the solution or the left sidebar navigation.

2. Click the title of the dataset to access the dataset overview and click Explore dataset.

3. Select the "Create insight" option to launch the insight editor.

4. Enter a name and description for the insight.

5. The insight editor will open from the right side.

6. The dataset you came from is already selected by default for your insight. You can also change it here.

From a solution

1. Navigate to a solution.

2. Create a new insight using one of the following options:

   • Click on the Create insight button on the Overview tab.

   • Or click on + New insight on the Insights tab.

3. Enter a suitable name and description for the insight.

4. The insight editor will open from the right side.

5. Select the dataset you want to use as the basis for your insight.

From the insights overview

1. Click on "Insights" in the left navigation bar.

2. Click the + New insight button.

3. Enter a suitable name and description for the insight.

4. The insight editor will open.

5. Select the dataset you want to use as the basis for your insight.

You can only use datasets for which you have either editor or owner permissions.

2

Join another dataset (optional)

1. After selecting your main dataset, click the + Join dataset button at the top of the Data tab.

2. Select the second dataset you want to combine with the main dataset.

3. The Polyteia Platform will attempt to automatically join them via matching column names (a natural join).

4. If the join key columns have different names, click the ··· button next to the second dataset name, then Add condition to manually define the column pair.

5. Use the dropdown menu to select your join type:

   • Inner Join, Left Join, Right Join, or Full Join.

6. After joining, you can select columns from both datasets for analysis.

Rename the join columns in your datasets to match so that automatic joins work seamlessly.

3

Select relevant columns

Via table header

1. Once the dataset (or joined datasets) is selected, the raw data table (joined data) will be displayed in the middle area with all columns

2. Click on the column header to add or remove columns for the analysis

3. You also have the option here to rename the column for this insight. To rename the column, simply click on the column name, enter the new name and click on the checkmark

4. You can also change the column order here

Via right navigation

1. Once the dataset (or joined datasets) is selected, the raw data table (joined data) will be displayed in the middle area with all columns

2. Click "+ Add column" to add columns to the analysis. The table in the middle will update to show only your selection

3. You also have the option to rename the column for this insight. To rename the column, simply select the text field next to the column name and enter the new name. This automatically updates the data preview table in the middle

4

Aggregate and group columns

1. To aggregate a column, select the · · · menu next to the column name.

2. Depending on the data type of the column, you can now select an aggregation function. For a detailed explanation, go to the Aggregation section.

3. Groupings allow you to summarize data by certain categories. The insight editor automatically groups by other columns that haven't been aggregated.

4. Each group is then displayed in the data table in the middle area of the editor with its associated aggregated values.

Enhanced interactive data exploration

The data table in the middle area provides powerful tools for exploring your data directly with enhanced functionality:

Interactive column headers:

1. Click column names to rename them inline with save/cancel options

2. Add/remove columns using left/right insertion buttons when available

3. Reorder columns by dragging or using move controls

4. Visual indicators show applied filters and current sort direction

Advanced sorting options:

1. Click the three dots (⋮) in any column header

2. Choose "Sort ascending", "Sort descending", or "No sort"

3. Segmented controls provide modern UI for sort selection

4. Sort state preservation maintains your settings when switching modes

Comprehensive filtering:

1. Click the three dots (⋮) in any column header

2. Select "Add filter" to access advanced filter options

3. Multiple filters per column - add as many as needed

4. Filter chips appear in headers showing active filters

5. Batch filter management - easily add, remove, or modify filters

Enhanced filter types by data:

• Text columns: equals, not equals, contains, starts with, ends with, is null, is not null

• Number columns: equals, not equals, greater than, less than, greater than or equal, less than or equal, between ranges

• Date columns: equals, before, after, between date ranges, in quarter, in year

Improved data interaction:

• Copy any cell value by clicking on it

• Type-aware cell display with proper formatting for dates, numbers, and text

• Null value indicators clearly show empty or missing data

• Popover details for long content that doesn't fit in cells

• Row limit controls - display up to 15,000 rows (default 100) with performance optimization

Performance features:

• Virtualization handles large datasets smoothly

• Sticky headers remain visible during scrolling

• Column resizing with draggable borders

• Responsive design adapts to screen size

This enhanced exploration environment lets you deeply understand your data patterns, quality, and structure before creating visualizations. Perfect for questions like "Show me only records from District 5 with high vaccination rates" or "Find all entries where the response time exceeded 10 minutes."

#### Make pivot tables

Think of a pivot table like organizing your closet by color and type of clothing. You can see all your blue shirts in one spot, all your red pants in another, and so on:

1. After picking your columns, turn on the "Make pivot table" switch.

2. Set up your table layout:

   • Rows: Click "Add row" to choose what goes down the left side (like listing all your shirt colors)

   • Columns: Click "Add column" to choose what goes across the top (like listing clothing types)

   • Values: Click "Add value" to choose what you want to count or add up in each box

3. The Polyteia software will automatically:

   • Add up number columns for you

   • Count non-number columns for you

   • Show you all your data without limits

4. Drag items around to change the order.

5. Your table will show you totals where each row and column meet - like seeing how many blue shirts you have.

Pivot tables are perfect when you want to compare things in two different ways at once. For more help, see the Pivot Tables guide.

5

Apply filters

Via table header

1. Click on the three dots (⋮) in any column

2. Select "Add filter" to access advanced filter options

3. Multiple filters per column - add as many as needed

4. Filter chips appear in headers under the three dots and show active filters

Via right navigation

1. Click the + button next to the filter category

2. Select the column to which the filter should be applied. You can select columns from your selection above and also columns from the (joined) dataset that are not visible in the data preview in the middle

3. Next to the selected filter column, you can now select filter options such as "equals" or "is greater than". For a detailed overview of all filter options, go to the Filter section

4. You can also aggregate a column and then filter based on the aggregation results by selecting the · · · menu on the filter column

5. Enter the value you want to filter by next to the filter option

6. The data table in the middle area will automatically update with the filtered data

6

Sort and limit rows

Via table header

1. Click on the three dots (⋮) in any column

2. Select "Sort ascending", "Sort descending", or "No sort"

Via right navigation

1. Click the + button next to the sorting category

2. Select the column to which the sorting should be applied. You can select columns from your selection above and also columns from the dataset that are not present in the middle area of the editor

3. Next to the selected sort column, you can now select sort options such as "ascending" or "descending"

4. You can also aggregate a column and then sort the results with the aggregation by selecting the · · · menu on the sort column

Enhanced row limits help optimize performance while showing more data:

• Default: 100 rows for quick loading

• Custom limits: Set up to 15,000 rows based on your analysis needs

• Performance protection: Hard cap prevents UI slowdown with large datasets

• SQL mode: Custom LIMIT clauses in your queries are preserved

7

Save, edit, delete, and export insights

In the top right corner, you'll find various actions you can perform.

1. Save: Save the insight.

2. Edit: Use the Edit button to edit the name and description of the insight.

3. Export: Export the insight.

4. Save as dataset: Save the results of the insight in a dataset.

5. Copy ID: Copy the unique ID of the insight.

6. Delete: Use the red trash button to delete an insight.

Don't forget to save changes after creating or editing the insight. Otherwise, they might be lost.

8

Create visualizations

1. Click on the Chart tab at the top right in the editor (next to the Data tab).

2. Select a chart type from the dropdown menu. Available options:

   • Table

   • Pie chart

   • Bar chart

   • Line chart

   • Single number

   • Map

3. Set up your chart by picking the right columns for what goes where in your chart.

4. For Table charts, you can switch between a regular table and a pivot table using the settings from Step 4.

5. Change the chart title and description if you want. It will use your insight name by default.

6. Look at the chart preview to see how your data will look.

Charts work best when you've already organized and summarized your data. Think of it like cleaning your desk before taking a photo - tidy data makes much prettier charts. Pivot tables are especially good for comparing different things side by side.