How This Documentation Is Organized

Overview

The Overview page is the front page of this documentation and provides a list with links to the respective subsection and some information on the estimations and descriptive statistics run.

Variables

The Variables page lists all variables found throught the data files found in the folder. Currently only Stata (.dta) and csv files are suported (and the latter are analysed using Stata as well). The variables are organised by their name and each name as a subpage that list detailed information about the respective variable in the relevant data file.

Files

The documentation can recognise a number of different file formats each get their own subsection in the Files page:

• Data
• Scripts
• Images
• Documents
• Other

Data

Recognised data files are examined for the varibales they contain and are linked to the script files where they are produced or used.

Scripts

Scripts are the central hub for the documentation that link datafiles and variables across inputs and outputs. Currently the script files are parsed for specific types of commands, including:

• System (changes to the logic and program environment)
• Manipulation (manipulation of variables)
• Input (commands that read data from files)
• Output (commands that produce files)
• Stat (commands that produce summary statistics and/or graphs)
• Estimations (commands that perform estimations)
• Call (commands that call other script files)

The classification of commands for Stata can be changed in the config/statacmdtype.txt configuration file. Statdoc tries to find the corresponding inputs and outputs for each script and lists them at the top of the page. You can help Statdoc in this task by using unique file names throughout the project. It is able to perform limited wildcard matching for filenames that include locals and globals if the remainer of the file name is sufficiently unique.

Scripts can contain documenting comments, which are marked with /** */ (not there are exactely two stars). These comments will be parsed for additional tags (e.g. @author) that will be added to the documentaion.

Statdoc also recognises when a log file is writen and will display the log file alongside the commands it has found (look for the the log button in the top right corner of the command list).

Images

Statdoc recognises image files and displays them if the relative path is intact (i.e. the statdoc folder must be in the root folder of the project).

Documents

Statdoc recognises Document files and can link them to script files (e.g. log files). In the future Statdoc might index the document file for more cross links (also see Tokens).

Other Files

Other Files list all other files currently not recognised/processed by Statdoc.

Tokens

The Index contains an alphabetic list of all n-gram tokens found throughout the directory. It is currently mainly limited to file names but might be expanded in the future. Then it would serve as the main index.

Prev/Next

These links point to the next or previous item (alphabetically) of a given type.

Frames/No Frames

These links show and hide the HTML frames. All pages are available with or without frames.

How to Make it Work Well

Statdoc tries hard to discover features of yout Statistics project by itself. There are a few things you can do to help though:

• General
• Use unique, descriptive and concise names for files and variables. Start names for related thins with the same string (e.g. same file with different data on the end).
• Scripts
• Use a documenting comment (starting with /** at the beginning of the file).
• Start your script by setting up a log file.
• If possible try to strictly seperate data manipulation and data analysis (differnt files).
• Use descriptive adn unique tokens for local variables.
• Avoid extensive abbreviation, use full variablenames.
• When using multiple input files in a loop, make sure there is a unique stub present that allows Statdoc to properly identify the files.
• Data
• Make sure data is properly labeled and in the appropriate format (double,float,int,str,...).
• Use xtset or tsset to identify your key variables.

How to Customise It

A lot of things are customisable for Statdoc. Starting with stylesheet.css, where variouse aspects of the display in html can be adjusted, e.g. the color of most elements. Furthermore all html pages are produced with the templates in the templates directory, this allows a lot of freedom in defining what is displayed in which order and where. In the same folder there are also a number of .do files that govern how information is extracted from Stata files. Finally the config folder holds some configuration files which allow customisations of how data is read.