Algorithm Documentation

Summary

This page deals with the specifics of how to document an algorithm. For a more general guide to the Mantid documentation system see Documentation Guide For Devs.

How to Document an Algorithm

Algorithm documentation is stored in two places.

  • The code (.cpp / .h / .py) files: For strings that are needed in the GUI for tooltips etc.

  • The .rst file: For all other documentation, including the algorithm description and usage examples.

The Code Files

The code files hold documentation in two important areas.

The summary method

The summary method should return a string (in plain text) that describes in a short sentence the purpose of the algorithm. This is used to provide a summary in the algorithm dialog box and in the algorithm documentation web page.

In C++: (This example uses the .h file, you could of course implement it in the .cpp file)

/// Algorithm's name for identification overriding a virtual method
const std::string name() const override { return "Rebin";}

///Summary of algorithms purpose
const std::string summary() const override {return "Rebins data with new X bin boundaries.";}

Property Descriptions

Each property declaration in the init method of the .cpp file should include a description. This is used as a tooltip, and in the ‘properties’ table on the algorithm’s documentation web page.

Note

The tooltip may only show the description up to the first dot. Be sure that the first sentence in the description gives enough information to understand the purpose of the property.

For example:

declareProperty(
new WorkspaceProperty<>("InputWorkspace", "", Direction::Input),
  "Workspace containing the input data");
declareProperty(
new WorkspaceProperty<>("OutputWorkspace","",Direction::Output),
  "The name to give the output workspace");
declareProperty(
new ArrayProperty<double>("Params", std::make_shared<RebinParamsValidator>()),
  "A comma separated list of first bin boundary, width, last bin boundary. Optionally "
  "this can be followed by a comma and more widths and last boundary pairs. "
  "Optionally this can also be a single number, which is the bin width. "
  "In this case, the boundary of binning will be determined by minimum and maximum TOF "
  "values among all events, or previous binning boundary, in case of event Workspace, or "
  "non-event Workspace, respectively. Negative width values indicate logarithmic binning. ");

Workflow algorithms

There should be a flow chart for workflow algorithms. See here on how to make one.

Algorithm Directives

We have defined several custom directives using the Sphinx extension framework. These are defined to generate the information from an algorithm that is not required to be hand written, i.e the properties table. Each .rst file that is documenting an algorithm should use these directives.

As the Description and Usage of an algorithm cannot be obtained automatically, it must be manually entered. The structure of an algorithm’s .rst is:

.. algorithm::

.. summary::

.. relatedalgorithms::

.. properties::

Description
-----------

The description of your algorithm here.

Usage
-----

A custom usage example.

.. categories::

.. sourcelink::
.. algorithm ::

This directive has several pieces of functionality, which includes:

  • A referable link is created for the algorithm. This allows other documentation pages create references to it (e.g. create links to it).

  • Insertion of the page title (which is the name of the algorithm, including the version).

  • Insertion of a screenshot of the algorithm’s dialog.

  • Insertion of the Table Of Contents.

.. summary::

The content of the summary method declared in your algorithm is output as HTML, for example, the following method is used in Rebin:

/// Summary of algorithms purpose
const std::string summary() const override {
   return "Rebins data with new X bin boundaries. For EventWorkspaces, you can very quickly rebin in-place by keeping the same output name and PreserveEvents=true.";
}
.. relatedalgorithms::

This directive obtains a list of related algorithms from the seeAlso and alias methods in the algorithm, for example, the following seeAlso is used in Rebin:

/// Algorithm's seeAlso
const std::vector<std::string> seeAlso() const override {
 return {"RebinToWorkspace", "Rebin2D",           "Rebunch",
         "Regroup",          "RebinByPulseTimes", "RebinByTimeAtSample"};
}
.. properties::

As mentioned above, it is critical that you include a description for the properties of your algorithm. This directive obtains all of the algorithm’s properties (set inside the algorithm’s init method) and outputs in a table format.

.. categories::

By default, this directive obtains the categories that were set in the categories method the algorithm. For example, in Rebin the category method is in the header and contains:

/// Algorithm's category for identification
const std::string category() const override {return "Transforms\\Rebin";}

When the HTML is generate a categories list is built that contains: Algorithms, Transforms and Rebin.

It is possible to add additional categories by passing the directive arguments, for example, the following would output the above categories, and also Example:

.. sourcelink::

This directive adds links to the algorithms source code.

Description

This section must be manually entered. The description is an extension of the summary, and must contain a detailed overview of the algorithm’s functionality.

Referencing Other Algorithms

Every algorithm and version has been marked with a tag that can be used to reference it from other documents. If you need to reference the latest version of an algorithm, e.g. DiffractionFocussing, then in Sphinx you type would type

:ref:`DiffractionFocussing <algm-DiffractionFocussing>`

If you need to reference a particular version then you would type

:ref:`DiffractionFocussing version 2 <algm-DiffractionFocussing-v2>`

where the first part outside the angle brackets defines the link text and the part inside the angle brackets references the tag name.

Usage

This section must be manually entered. The usage is a ‘code’ example of the algorithm in use. The .. testcode:: directive must be used to verify the usage code you entered works correctly. See here for more information on how to write usage examples.

Building the Documentation

One can update the documentation for a particular algorithm after changes have been introduced into the corresponding documentation file. Assuming you are in the build directory and want to update the documentation for Rebin:

PYTHONPATH=/path/to/build/bin python3 -m sphinx <sourcedir>/docs/source docs <sourcedir>/docs/source/algorithms/Rebin-v1.rst