Application Profiler

The Application Profiler is used to gather information about analysis files. Given a set of files, local or from the library, the tool loads each file and outputs a number of data tables in STDF files (e.g., one with one row per analysis file, one with one row per data table, and one with one row per visualization). The Application Profiler is a useful tool before or after applying a hotfix or upgrading to a major release of Spotfire. As the output from the tool can be examined, it allows you to monitor whether any visualizations or other parts of the analysis appear or behave differently over time (e.g., how many visualizations do the analyses have, has the total rendering time changed over the last three months, has there been any changes in the time it takes to load data in a specific analysis).

Note: To get access to the Application Profiler, Spotfire Diagnostics, as well as the license feature called Application Profiler below it, must be enabled in Administration Manager.

To reach the Application Profiler dialog:

Select Tools > Diagnostics > Application Profiler.

General information

The Application Profiler allows you to customize paths, options, and scripts, and save them locally to your computer as a configuration, ready to be loaded into the tool again.

Option	Description
Load Configuration	Allows you to open a previously saved configuration.
Save Configuration	Saves a configuration locally on your computer.
Reset Configuration	Clears all paths, options, and scripts in the dialog.
Included paths	Lists added files or folders from either your local computer or from the library.
Add	Opens a dialog allowing you to select a folder or a file, local or from the library.
Remove	Removes the selected folders or files from the list.

Extra test options

The tool can optionally perform a set of predefined tests. If no options are selected, the tool only gathers basic information.

Option	Description
Extra test options	List of available options for testing and rendering. Most of these options perform an implicit undo after each operation.
Render original visualizations	'Original' refers to the visualization the way it looks when opening the analysis for the first time (e.g., without applied bookmarks or filtering). In addition to writing out basic information about the visualization, this option adds columns for visualization errors, rendering issues, rendering time, whether the visualization configuration is valid, and the rendered image.
Render using obsolete pre-7.12 API	The rendering code for visualizations was updated in Spotfire 7.12. If you have run Application Profiler earlier, you might see small differences in images due to the changed rendering when running an updated version. If it is important not to get any differences compared to your previous tests, you can try rendering using the old API.
Interact with visualizations	Interacts with visualizations using e.g., marking or highlighting.
Perform filtering	Performs filtering and renders all visualizations for all combinations of filter type and data type.
Apply bookmarks	Applies all bookmarks found in the analysis, and for each bookmark, renders all visualizations.
Remove/rename columns	For each column in each data table, the column is removed and all visualizations are rendered. After that, each column in each data table is renamed and all visualizations are rendered again.
Remove/rename data tables	For each data table, the table is removed, and all visualizations are rendered. After that, each data table is renamed and all visualizations are rendered again.
Create new visualizations	Creates one new visualization of each type.
Save and Reload	After performing all available test options on an analysis, the file is reloaded, saved in the current file format, and tested again.
Time-out per analysis (in minutes)	A time-out is applied once for the original version of the analysis, and is initiated a second time for the new version when using Save and Reload.
Max number of test threads	The maximum number of concurrent threads, one analysis per thread.
Allow data function scripts execution	Selecting this check box means that you want to run data functions which are built into the analysis. Leaving the check box without a check mark may result in for example a data function error, a version related error or an empty visualization, but might prevent unwanted scripts to affect the system.
Edit	Allows you to set the test option to run all or a limited number (e.g., filters, bookmarks, columns and data tables).

Scripts

The tool allows you to load customized scripts in order to highlight and present specific data in a document, data table or a visual.

Option	Description
Scripts	Supports custom scripts that can be executed once for each document, data table and visual.
Document	Runs a script for each document. A variable called Document will be available.
DataTable	Runs a script for each data table. A variable called DataTable will be available. Note: Because the variable name is the same as the object type, the script cannot contain structures like: from Spotfire.Dxp.Data import * This will stop the DataTable object from having a Name property and no outputs will be shown. Instead, when you need to import from Spotfire.Dxp.Data, do it using specific names rather than the wildcard.
Visual	Runs a script for each visual. A variable called Visual will be available.
Script to run once for each ...	Shows the current script to run.
Available OutputColumns for each ...	Shows the current script used to configure additional output columns for the document, data table, or visualization output data tables.

Example

The following IronPython scripts can be used to gather information about bar orientation in all the bar charts in a set of DXP files. The output column script defines a column called BarChartOientation with data type string, and the main script gathers information by outputting values to that column.

To create the script in IronPython for defining available output columns:

Open a new document in a text editor of your choice.
Type the following script:

Note: For defining available output columns, you always have to add to OutputColumnDataTypes

from System import Tuple, String
from Spotfire.Dxp.Data import DataType

OutputColumnDataTypes.Add(Tuple.Create[String,DataType]("BarChartOrientation", DataType.String))

Save the document locally to your computer with the file extension suffix .py.
Load in Available OutputColumns for each Visual.

To create the main script in IronPython to run once for each visual:

Open a new document in a text editor of your choice.

Type the following script:

import Spotfire.Dxp.Application.Visuals.BarChartOrientation as barChartOrientation
from Spotfire.Dxp.Application.Visuals import BarChart
from Spotfire.Dxp.Application.Visuals import VisualTypeIdentifiers

if Visual.TypeId == VisualTypeIdentifiers.BarChart:
           if Visual.As[BarChart]().Orientation == barChartOrientation.Horizontal:
                       OutputColumns["BarChartOrientation"] = "HORIZONTAL"
           else:
                       if Visual.As[BarChart]().Orientation == barChartOrientation.Vertical:
                                   OutputColumns["BarChartOrientation"] = "VERTICAL"

Save the document locally to your computer with the file extension suffix .py.
Load in Script to run once for each Visual.

Executing the Application Profiler

Before executing the Application Profiler, you need to specify where to save the output data tables and give a title prefix to the output data tables.

Option	Description
Save output tables to	Specifies where to save the output data tables.
Prefix	A title prefix for the output data tables which are saved locally to your computer.
Start	Starts the test by downloading and running tests on added files or folders.
Show Analysis	Provides a quick shortcut to create an analysis from where the output tables are saved.