Summary: Pipette Analysis Charting allows scientists to visually explore data in a web-based interface that is both powerful and intuitive. This article provides details about how to create charts.
Clicking a link will scroll the page to the relevant section.
- Getting Started
- Data Types
- Chart Types
- Configuring a Chart
- Filtering Data
- Creating Multiple Charts
- Chart Templates
- Exporting Selections
- Keyboard Shortcuts
- Pipeline Pilot Components
- Data Modes
- Integration with Other Systems
To explore your first data set using Pipette Analysis Charting, work with data that has a variety of numeric and string properties.
Drag your data set to the Charting output in the Pipette Analysis Pipelining application.
The Pipette Analysis Charting application will load, and several interconnected histograms showing the first eight properties in your dataset are displayed.
The following options are available:
- Create a new chart and start exploring your data
- Load a pre-defined template Chart Templates that shows your data in one or more charts.
For this example, we will create our own chart.
To create a chart:
- Select Configure Chart (top left) to open to chart configuration dialog.
- Click Change Plot Type to see all available charts.
- Select XY Plot.
- Set up each dimension on the XY Plot by selecting it and then choosing from the list of properties on the right.
- Click OK to display your chart.
- You can add more charts, select and export data, and modify your chart.
- The X and Y axes are required, but radius and color are optional.
- The list of properties is automatically filtered and only those that are valid for the selected dimension are displayed.
When data is imported in Pipette Analysis Charting, it is automatically assigned a type. This determines how the data can be used in the charts. Indicators for the different chart types are displayed in a configuration dialog, as illustrated in the following example.
|Whole Numbers||1, 4, 100||123|
|Real Numbers||3.14, -2.5||123|
Complex data structures, such as arrays of data, cannot be used in the Pipette Analysis Charting. To chart this type of data, aggregate the values (often mean, median, min, or max) before sending the data into Pipette Analysis Charting.
All data types can be marked as categorical. This happens when they have a restricted set of values across the data set. If a property has less than 20 values, use it to trellis XY plots, display it in a histogram (even if it is not numeric), or color points on 3D scatter plots, among many other uses. Categorical strings have a special indicator in the chart configuration dialog (a,b,c).
Numeric values without a range, where the value of a property is the same for every record in the dataset, are considered categorical, and cannot be used in situations where a numeric range is expected, (e.g., the X axis of an XY plot).
A range of chart types are available in Pipette Analysis Charting. For details, see Pipette Analysis Reference – Chart Types.
Configuring a Chart
Use the Configure Chart feature on the Configuration tab to set up a chart. It displays the main configuration window shown below.
You can change the property on an axis by selecting the axis on the chart itself.
Some charts have options to control their configuration. These features displayed in the Configuration tab as a series of buttons.
You can view other charts that can display the current data in different ways by selecting Alternative Charts. For example, you can switch between stacked histograms, a radar plot, a box plot and a matrix correlation plot.
Use the Filtering tab to reduce the data set to a subset of interest.
Note: Filters apply to all currently displayed charts.
Creating Multiple Charts
Use the Add Chart buttons on the right and bottom of the screen to create multiple charts.
The charts are interconnected, meaning that data selected on one chart can be highlighted on other charts, provided that the other charts can show the data in a meaningful way. This feature is useful for exploring relationships between data.
The templates tab displays a list of chart templates that you can apply to the current data. This allows you to quickly display your favorite charts and dashboards.
The list is sorted and colored to indicate if the template applies to the current data.
To see if the template uses any properties that are not available on the current dataset, hover over the template and inspect the tooltip. The impact of a missing property will depend on the dashboard – it could mean one chart on a dashboard is not shown, or points on an XY plot do not vary in color.
Templates have two general scopes. This is indicated by the icon on each template:
In ScienceCloud, each organization has its own shared templates. If you are member of multiple organizations, you can use shared templates from each one.
The current chart configuration can be saved as a template and applied to another data set in the future. From the Filter tab, use the Save Template button for this purpose.
When saving a template, you need to provide a name, summary, and description. The summary is important, because it is used with the name when users search for templates. This makes it an excellent location for useful search keywords.
All users have the ability to save user-scope templates, but administrators have the ability to control who can create and update shared templates. For details, see Pipette Analysis On-Prem Administration Guide.
All currently visible charts and their settings are saved when a template is saved.
Any data selection from a chart can be exported to a range of outputs. This is a useful way to send a subset of data back to Pipette Analysis Pipelining, to reports, or other systems.
These outputs match those found in Pipette Analysis Pipelining. In Pipette Analysis Charting, they are available from the Outputs tab.
The list of outputs is ordered by how frequently they are used.
Pipette Analysis Charting has keyboard shortcuts for common tasks. For details, see Pipette Analysis/Charting Reference - Keyboard Shortcuts.
Pipeline Pilot Components
The following Pipeline Pilot components work with Pipette Analysis Charting:
- Pipette Analysis Charting Viewer – Sends the current data you are working with in Pipeline Pilot to Pipette Analysis Charting.
- Pipette Analysis Dashboard Builder – Serializes the name and parameters of the current protocol to the charting URL. This URL can be shared or bookmarked to allow users to get back to the dashboard. Each time the dashboard is loaded, the underlying protocol runs to provide the latest data.
Pipette Analysis Charting has two modes depending on the data size. What constitutes "large" data depends on several factors, including the web browser, type of device (tablet or PC), number of properties per data record, and the total number of data records.
Client-Side Data Mode
This is the primary mode for Pipette Analysis Charting. All data is loaded, aggregated and presented in the web browser. This gives great performance and eliminates any network delays after the data is loaded.
Server Side Data Mode
When data volumes grow large in size, the server does not try to send all of the data to the client. Instead, it is aggregated on the server and sent to the web browser. Some types of plots, such as XY plots, are not available for large data volumes. Instead, more useful and performant plots, such as the Surface Plot with Outliers, are used.
Example: The surface plot with outliers allows viewing of individual outliers, but presents the bulk of data as a surface chart, rather than an indistinguishable mass of points.
Integration with Other Systems
You can embed Pipette Analysis Charting in other applications or launch it from your application with a preconfigured dashboard. For details, see the Pipette Analysis Charting Integration Guide.