Getting Started

For those who want to explore their data visually, Visual Explorer is a more flexible and powerful alternative to our “fixed view” Quick Charts. Inspired by the layered Grammar of Graphics, Visual Explorer provides an environment where you can create and configure your visualizations based on a ruleset—a grammar, if you will. Just like how in English the word “grammar” has become synonymous with rules for the construction of readily interpretable sentences, Visual Explorer is built upon a framework that provides rules to construct readily interpretable graphics.

Consequently, there are a few key differences between the workflow within the Visual Explorer and Quick Charts environments:

1. The first question to be answered:

   • Quick Charts: what chart type do you want?
   • Visual Explorer: what features of your data do you want to examine?

2. The iteration process:

   • Quick Charts: arrive with a starting graphic type in mind, then rebuild through until you arrive at something you like
   • Visual Explorer: iterate toward your end visualization by using modular, independent components

As a general principle, you can think of Columns as what data you want to display horizontally (similar to an x-axis), Rows as what data you want to display vertically (similar to a y-axis), and Layers as how you want to display your data.

Building a Simple Visualization

Adding Dimensions and Measures

If you have a multi-dimensional dataset, adding a Dimension to your Columns or Rows is a powerful starting point to begin slicing your data. Doing so groups your visualization by categorical values, which helps you break down your data to better understand certain trends.

From there, you can drag and drop a Measure to your Rows or Columns. Measures are automatically aggregated by default when added to visualizations.

You can also choose to manually disaggregate Measures, if you want to plot specific data points on a continuous range. This is useful for numeric data that’s already meaningful without any aggregation.

Multiple Measures

From here, you can continue to add additional Measures to your visualization if you wish. By default, every measure is plotted in its own pane because each measure introduces a new layer.

Check out our Visual Explorer field guide for a walkthrough on how to create visualizations in the Visual Explorer. There you can find possible visualizations you can build as starting points for your exploratory analysis.

Introduce Groupings with Facets

Once you add multiple Dimensions to your Columns or Rows, your visualization will begin to nest and display facets. Nesting is a way to compare within a dimension’s values (rather than against them) and helps break down metrics across dimensions within the same plot. Think of this as equivalent to GROUP BY with multiple fields in SQL.

The new dimension divides the visualization into 10 panes, one for each combination of YEAR(order_created_date) and region_name.

Using Layers to Add Depth

Layers are an integral part of visual analysis because they allow you to add further context and detail to your visualizations. You can change the Layer properties of an individual measure or all measures in your visualization.

Mark Types

By default, the Mark Type of your visualization will be set to Automatic — based on your configuration, Visual Explorer will try to figure out what is the optimal mark to display. But of course, you can opt to set and change the Mark Type of each layer. Click the dropdown menu to select how you want to plot your measure.

In this example, we’re actually changing the type for the specific measure SUM(order_total_amt_usd).

Color

Dragging a Dimension to the Color channel will slice your Measure, where each value in your Dimension will be represented with a different color in your chosen color palette. Think of this as equivalent to a GROUP BY in SQL.

You can also drag and drop a Measure to the Color channel. This will produce a color gradient with the minimum and maximum values of that Measure based on the grouping in your visualization.

Size

You can add drag a continuous numeric Measure to the Size channel, so that the width of every mark in that particular layer will now reflect the value of that measure. This is commonly used in bubble charts to provide additional information about data points relative to each other.

Text

You can drag a field to the Text channel to display text on your visualization. This will display a text data value for every mark in that particular layer. You can use this to specify what labels to display in your

Detail

Like Color, when you drop a dimension onto the Detail dropzone, the marks in your visualization will be separated according to the values within that dimension. But unlike dropping a dimension on Rows or Columns, adding fields to Detail is a way to show more data without changing the table structure. However, they will appear in tooltips upon hover.

The grouping in Detail will also be factored into your calculations when you use window functions or Quick Calculations.

Building Visualizations with Multiple Measures

Measure Names and Measure Values are always available in the fields list. Unlike the other fields, they do not directly come from your dataset but are rather provided by the Visual Explorer for you to define a group of measures in your visualization.

• Measure Values contains the values of all the measures in the Measure Values shelf, collected into a single field with continuous values.
• Measure Names contains the names of all the measures in the Measure Values shelf, collected into a single field with discrete values.

By default, Measure Names and Measure Values will be empty variables for you to fill.

The combination of these two fields allow you to build certain types of views that involve multiple measures. As you’ll begin to see in some our our example configurations, Measure Names and Measure Values are integral pieces to build certain visualizations.

The Measure Values Shelf

Once you add Measure Values anywhere in your visualization configuration, a Measure Values shelf will show up for you to add your measures.

Unlike adding measures directly to Columns or Rows, this technique plots all measures in the Values dropzone in the same view.

Multiple Measures in a View

There are several ways to graph multiple measures in one view:

1. Create one axis for each measure
2. Blend two measures to share on axis
3. Set to measures on a dual axis

One Axis for Each Measure

By default, each measure gets its own axis when you add measures directly to either Rows or Columns.

• Adding a measure to Columns will create a new axis along the x-axis
• Adding a measure to Rows will create a new axis along the y-axis

In the example below, we have one measure in Rows and another measure in Columns. Compare that to the resulting visualization.

Blending Axes

If you want multiple measures to be in the same pane and axis, you’ll want to leverage Measure Values and Measure Names.

1. Drag and drop Measure Values to either Rows or Columns—whichever axis you want your measures to be on.

2. You’ll now see that a new Measure Values shelf has appeared. This is where you will add the measures you wish to include in your visualization.

3. You’ll want to add Measure Names to your visualization to ensure that you can distinguish between your measures. You can place it either on a Row or Column or in a layer channel like Color.

Dual Axes

Lastly, you can compare two measures on the same pane but different axes by creating visualization with a dual axis.

1. Drag and drop the at least two measures you want to graph on either Rows or Columns.
2. Click on a field that you’d like to include in your dual axis. In the context menu, you should see an option to join this field and the field above on a dual axis. If it’s the outermost field and there’s no field above it, then you will not see the dual axis option.

3. Lastly, add Measure Names to the Color dropzone if you wish to be able to visually distinguish between the two measures.

Data Limits within Visual Explorer

There are no limits to the amount of data you can pull into Visual Explorer; instead, that will be determined by your Helix tier.

However, there are browser limitations to what you can visualize. Visual Explorer will plot up to:

• 250000 individual data points (for example, in a scatter plot)
• 16000 facets in a pivot table or faceted chart
• 3000 series (lines or bars)

Dealing with High Cardinality

If you are working with datasets/fields that surpass browser limitations, there are a couple of actions you can take to render your visualizations:

• Aggregate a field: When you add a field that exceeds one of the browser limitations, you can aggregate your field with COUNT
• Applying a filter: You can also apply a list, top N, or bottom N filter to see a subset of your data and render your visualization.
• Turning off auto-update: You can switch to manual mode and delay all the updates when making changes to a visualization. In manual mode, a pop-up will appear and count all the unapplied changes. When you are done with your changes, hit “Apply changes” and the visualization will render.

Filters

Applying Top/Bottom N Filters

Top/bottom N filters enable you to see the highest or the lowest categorical levels ranked by aggregating a given measure. To apply a top or bottom N filter to your visualization:

1. Drag and drop your dimension (text, date, and numeric fields) into the filter section.
2. Select the “Top/Bottom N” method in the filter modal and choose what you want to define along with the number of values.
3. Select the field you want to filter by.
4. Choose the aggregation type you want for your filter.
5. Apply the filter and close.

Notes:

• Top/bottom N filters will always be applied first no matter the other filters that are added.
• In order to create a top or bottom N filter with no aggregation, you will need to select the same field you created with the by-field filter (see image below for an example).
• Currently top/bottom N filters do not support ties

Sorting

Quick Sort

In the Toolbar, you have the ability to leverage our Quick Sort feature to sort your innermost discrete, categorical data by the outermost measure in either descending or ascending order.

However, you also have the option to define a more granular sorting behavior. When you can click on a discrete field in your visualization to open its context menu, you will see the ability to open the sort dialogue.

From there, you can specify exactly how you’d like to sort that discrete field. We currently support two sorting methods:

1. By data value
2. By field
3. By manual
4. By nested

Sorting by data values

This sorting method looks at the values within the field you’re trying to sort by (e.g., Midwest, Northeast, Southeast, West within the region_name field) and sorts them either in ascending or descending order.

• For numeric fields, this refers to in order of smallest to largest or vice versa
• For string fields, this refers to alphabetical order
• For date fields, this refers to chronological order

Sorting by field

This sorting method allows you to sort a discrete field in your visualization—it could be in Columns, Rows, or one of your Layer channels—by another continuous, aggregated value. The latter field does not need to be in your chart configuration.

By-field is a non-nested sorting method that considers the total value across all panes and thus will have the same order of values across all panes.

Manual Sorting

The manual sort feature gives users the ability to sort a domain of items in whichever order they choose by allowing them to create a specific order through dragging and dropping values into a customized order.  Any items for the given field that have not been manually sorted with appear after the sorted items, in ascending order.

Nested Sorting

Nested sorting allows values to be sorted independently within each facet. To use nested sorting, the field you are sorting has to be under (or nested below) another field.

By field sorting and nested sorting are similar in that they allow a category to be sorted by a given variable. But while by field sorting disregards the nesting structure of faceted charts, nested sorting works within the constraints of that structure. Independent sorting is applied to the elements on each of the innermost facets.

Clearing your sort

You can tell when a sort has been applied to your visualization when you see this sort icon in the pills of any one of your discrete fields. At any point, you can opt to clear the sorts you’ve applied by either clicking on the field itself to open the context menu:

or by broadly clearing all sorts via the Toolbar:

Formatting Your Visualization

Before sharing your insights with a broader audience, you may want to format your visualization. The Visual Explorer gives you more granular control over your visualization than Quick Charts.

Anatomy of a Visualization

To understand how Formatting works in Visual Explorer, we first need to explain how we think about the parts that make up a visualization.

• Axis: Axes are created when you place a continuous field on Rows or Columns.

  • Axis Titles are the names of your axes.
  • Axes Values are the data values within an axis.

• Header: Headers are created when you place a discrete field on Rows or Columns.

  • Header Titles are the names of your headers.
  • Header Values are the data values within a header.
• Pane: Panes are formed when fields on Rows and Columns intersect. A visualization can be consisted of several panes.
• Legend: Legends are keys of the chart's data series to help you understand the visual representation of your data series, usually via Color or Size.

Formatting

When you click into the Format tab, you’ll notice that you’ll first be asked what field you’d like to format.

Depending on your selection, the relevant formatting configurations will then appear.

Axes

If you select a field that’s continuous, you’ll see that you’ll be able to format its corresponding axis. Note that whatever changes you make on the axes will not affect the contents in the pane.

• Title: You can toggle on/off the axis title or change the name of the axis

• Range: You can also set the axis range here. You have two options:

  • Independent: By clicking on the Independent checkbox, you are setting all the panes that use this axis to be independent of one another. That is, if you have multiple panes (i.e. your measure is nested underneath a dimension), the range of each pane should be determined individually.
  • Fixed: If you decide instead to set either a fixed minimum or maximum, you are opting for a fixed axis range. This means that every pane will share the same axis range, to be determined based on the overall minimum and overall maximum across all panes.
• Number format: You have the ability to change the formatting of your axis labels to best fit your data—decimal places, currency, percentage, etc. Note that this change will only be reflected in your axes. If you wish to change the numeric values within the pane, you’ll have to do so separately in the Pane section.

Headers

If you select a field that’s discrete, you’ll see that you’ll be able to format its corresponding header. A field can only have either an axis or a header—it cannot have both.

• Title: You can toggle on/off the header title or change the header title entirely. However, please note that whatever changes you make on the axes will not affect the contents in the pane.
• Show/Hide Values: You can also toggle on/off the header values.
• Format: Depending on the datatype of your discrete field (e.g. date part vs. string), you’ll see corresponding formatting options for your headers.

Panes

Whether a visualization contains an Axis vs. a Header depends on whether the field is continuous vs. discrete, but all visualizations will have a pane if there is an intersection of at least two fields. Any formatting changes you want to be reflected on your Labels or Tooltips should be made in the Pane section, as these elements all exist within the Pane.

The Visual Explorer will always allow you to configure the Pane, even if the field selected doesn’t currently render a pane in your visualization. This is so that if you later reconfigure your visualization, your work will be saved.

Legends

For fields that are referenced in your visualization’s legend, you will see a Legend section where you can opt to show/hide it, choose its positioning on your visualization, or rename it.

Manually resize columns

You can manually resize the widths of your column headers or axes:

1. Place your cursor at the right of the label of the column you wish to adjust
2. When you see the resize cursor, click and drag the border left or right

Quick Calculations

Overview

Quick Calculations allow you to quickly apply an analytic calculation over your data in your visualization. In addition to the calculation type, you can also specify the level and direction at which you wish to calculate for any aggregated, continuous field after you applied a Quick calculation to a measure.

You can access our Quick Calculations by clicking on the pill of any aggregated, continuous field.

Calculation Types

The calculation types that we currently support are:

• Percent Difference: Calculates the percent difference between the current value and another specified value for each mark in the visualization. In addition to the usual level and direction, you can also specify which value you’d like to anchor the difference calculation relative to.

• Difference: Calculates the difference between the current value and another specified value for each mark in the visualization. In addition to the usual level and direction, you can also specify which value you’d like to anchor the difference calculation relative to.

• Percent of Total: Calculates the value as a percentage of all values within a specified window in the visualization

Customizing Your Quick Calculation

In addition to what you want to calculate, you can also specify the direction (e.g. across vs. down) and the level (e.g. table vs. cell) you wish to calculate over.

• Table Across: Calculates across (left to right) the entire table and restarts after every partition

• Table Down: Calculates down (up to bottom) the entire table and restarts after every partition

• Table Across-Down: Calculates across (left to right) the table but does not restart after every partition

• Table Down-Across: Calculates down (up to bottom) the length of the table but does not restart after every partition

• Pane Across: Calculates across (left to right) the pane and restarts after every partition

• Pane Down: Calculates down (up to bottom) the pane and restarts after every partition

• Pane Across-Down: Calculates across (left to right) the pane but does not restart after every partition

• Pane Down-Across: Calculates down (up to bottom) the pane but does not restart after every partition
• Cell: Calculates within a single cell

FAQs

How do I customize tooltips?

You can add a field to the Detail layer channel. The field values will appear in the tooltip without affecting how the visualization looks.

Adding a dimension will separate the marks in your visualization to the values within that dimension, while adding a measure will not because it will be pivoted along the same dimensions.

Can I convert a Quick Chart to the Visual Explorer environment?

If you’re unsure how to get started in the Visual Explorer, a great way is to begin your visualization using our Quick Charts and then convert your work over to Visual Explorer by opening the Environment Switch. We support all Quick Chart conversions except for Tables and Big Values. We do not support converting Visual Explorer visualizations back to Quick Charts, though you can use the Back and Forward buttons in the Chart Editor.

How do I assign a specific color to a specific value in my visualization?

The Visual Explorer gives you more granular control over parts of your visualization than Quick Charts do. One of those controls is colors.

No Field in Color Channel

When you don’t have any field in your Color channel and you click on the Color icon to edit your colors, choosing any color will directly apply to that series.

Discrete Field in Color Channel

When you have a discrete field in your Color channel and you click on the Color icon to edit your colors, you’ll see your palette selection on the left hand side and the data values within your discrete field on the right hand side.

On the left, you can change and apply different palettes to your visualization. On the right, you can select individual data values and pick the specific color from any palette that you wish to apply to that selected data value.

Continuous Field in Color Channel

When you have a continuous, field in your Color channel and you click on the Color icon to edit your colors, you’ll see the ability to select the anchor color that will generate a continuous palette.

How do I sort in Visual Explorer?

In the Toolbar, you have the ability to leverage our Quick Sort feature to sort your innermost discrete, categorical data by the outermost measure in either descending or ascending order.

Can I apply conditional formatting to my pivot table?

You can add color to the text or background in your pivot table based on some specifications by leveraging the color channel.

Discrete Field in Color Channel

You can create a calculated field with IF or CASE statement to form a categorical groupings that you can then apply to your pivot table.

Continuous Field in Color Channel

You can add a continuous, aggregated field to the Color channel. This will color the values in your pivot table based on the the cell’s value according to that field. The smaller the value, the lighter the color; the larger the value, the darker the color.

Can I have multiple fields in a Layer channel?

Visual Explorer currently only supports one field per channel in a given Layer.

Can I display Grand Totals in my chart?

You can toggle on and off either or both Column Grand Totals and Row Grand Totals using the Toolbar, like so:

By default, Grand Totals will take on the same aggregation method as the values you are totaling over—for example, it will sum all the values if your measure is SUM(field) or it will average across all the values if you measure is AVG(field).