gmoTree Codebooks

Patricia F. Zauchner

2023-09-21 (updated: 2024-12-06)

Overview

The gmoTree package provides tools for generating codebooks based on data from the oTree (Chen et al., 2016) framework with the function codebook(). This vignette introduces its key features and functionalities.

Disclaimer: This code has been tested with both custom implementations and official oTree sample codes.1 However, due to the complexity of oTree setups, it may not address all variations. Please review your output carefully, and feel free to open a GitHub issue if you encounter any problems.

Running the code

In its simplest form, the function requires only the path to your oTree code. The following code reads the contents of your oTree file, saves a file containing your codebook in your working directory, and returns a list named cbook, containing the extracted information.

cbook <- codebook("D:/pathtoyourproject/oTree_codes")

Codebook structure and output

The output is structured with clear headings:

Image showing an example of a heading structure of a PDF file with Level 1 and Level 2 headings

The example below shows the output for the cournot app, which includes all Constants, Group, and Player variables. If a class contains no information, the heading will still appear, along with the message “Empty class.”

Image showing the rankaversion app information in the PDF output

Use output = "list" to bypass file creation. This option returns all the information in list format, which can then be used in R Markdown to design a custom codebook. Below is an example of how such a list output looks like. (The argument doc_info = FALSE suppresses messages about missing variable documentation.)

# Create the codebook list
cbook <- codebook(path = system.file(path = "extdata/ocode_new", 
                                     package = "gmoTree"), 
                  output = "list", 
                  doc_info = FALSE)


str(cbook, 1)  # Show all apps 
## List of 16
##  $ settings            :List of 10
##  $ bargaining          :List of 4
##  $ bertrand            :List of 4
##  $ common_value_auction:List of 4
##  $ cournot             :List of 4
##  $ dictator            :List of 4
##  $ guess_two_thirds    :List of 4
##  $ matching_pennies    :List of 4
##  $ payment_info        :List of 4
##  $ prisoner            :List of 4
##  $ public_goods_simple :List of 3
##  $ survey              :List of 3
##  $ traveler_dilemma    :List of 4
##  $ trust               :List of 3
##  $ trust_simple        :List of 3
##  $ volunteer_dilemma   :List of 4
print(cbook$cournot)  # Only show cournot app
## $doc
## [1] "In Cournot competition, firms simultaneously decide the units of products to manufacture. The unit selling price depends on the total units produced. In this implementation, there are 2 firms competing for 1 period."
## 
## $Constants
## $Constants$NAME_IN_URL
## [1] "cournot"
## 
## $Constants$PLAYERS_PER_GROUP
## [1] 2
## 
## $Constants$NUM_ROUNDS
## [1] 1
## 
## $Constants$TOTAL_CAPACITY
## [1] 60
## 
## $Constants$MAX_UNITS_PER_PLAYER
## [1] "int(60 / 2)"
## 
## 
## $Group
## $Group$unit_price
## $Group$unit_price$noargs
## [1] TRUE
## 
## $Group$unit_price$field
## [1] "CurrencyField"
## 
## 
## $Group$total_units
## $Group$total_units$noargs
## [1] FALSE
## 
## $Group$total_units$doc
## [1] "Total units produced by all players"
## 
## $Group$total_units$field
## [1] "IntegerField"
## 
## 
## 
## $Player
## $Player$units
## $Player$units$noargs
## [1] "FALSE"
## 
## $Player$units$min
## [1] "0"
## 
## $Player$units$max
## [1] "int(60 / 2)"
## 
## $Player$units$doc
## [1] "Quantity of units to produce"
## 
## $Player$units$label
## [1] "How many units will you produce (from 0 to 30)?"
## 
## $Player$units$field
## [1] "IntegerField"

Choosing what to show

Apps

Per default, all apps are included in the codebook. To generate a codebook for only one app, use the app argument. For example, to generate a codebook for the cournot app:

cbook <- codebook(
  path = system.file("extdata/ocode_new",
                     package = "gmoTree"),
  fsource = "init",
  output = "list",
  doc_info = FALSE,
  app = "cournot")

str(cbook)
## List of 2
##  $ settings:List of 10
##   ..$ variable                : num 66
##   ..$ PARTICIPANT_FIELDS      : list()
##   ..$ SESSION_FIELDS          : list()
##   ..$ LANGUAGE_CODE           : chr "en"
##   ..$ REAL_WORLD_CURRENCY_CODE: chr "USD"
##   ..$ USE_POINTS              : chr "True"
##   ..$ ADMIN_USERNAME          : chr "admin"
##   ..$ ADMIN_PASSWORD          : chr "environ.get('OTREE_ADMIN_PASSWORD')"
##   ..$ DEMO_PAGE_INTRO_HTML    : chr "Here are some oTree games."
##   ..$ SECRET_KEY              : chr "9356668094877"
##  $ cournot :List of 4
##   ..$ doc      : chr "In Cournot competition, firms simultaneously decide the units of products to manufacture. The unit selling pric"| __truncated__
##   ..$ Constants:List of 5
##   .. ..$ NAME_IN_URL         : chr "cournot"
##   .. ..$ PLAYERS_PER_GROUP   : num 2
##   .. ..$ NUM_ROUNDS          : num 1
##   .. ..$ TOTAL_CAPACITY      : num 60
##   .. ..$ MAX_UNITS_PER_PLAYER: chr "int(60 / 2)"
##   ..$ Group    :List of 2
##   .. ..$ unit_price :List of 2
##   .. .. ..$ noargs: logi TRUE
##   .. .. ..$ field : chr "CurrencyField"
##   .. ..$ total_units:List of 3
##   .. .. ..$ noargs: logi FALSE
##   .. .. ..$ doc   : chr "Total units produced by all players"
##   .. .. ..$ field : chr "IntegerField"
##   ..$ Player   :List of 1
##   .. ..$ units:List of 6
##   .. .. ..$ noargs: chr "FALSE"
##   .. .. ..$ min   : chr "0"
##   .. .. ..$ max   : chr "int(60 / 2)"
##   .. .. ..$ doc   : chr "Quantity of units to produce"
##   .. .. ..$ label : chr "How many units will you produce (from 0 to 30)?"
##   .. .. ..$ field : chr "IntegerField"

The code above returns a list with the app-specific information. You can also generate codebooks for multiple apps by specifying them in the app argument.

cbook <- codebook(
  path = system.file("extdata/ocode_new",
                     package = "gmoTree"),
  fsource = "init",
  output = "list",
  doc_info = FALSE,
  app = c("cournot", "dictator"))

str(cbook, 1)
## List of 3
##  $ settings:List of 10
##  $ cournot :List of 4
##  $ dictator:List of 4

You can also choose all but remove one or more apps. Here is an example that removes the app prisoner:

cbook <- codebook(
  path = system.file("extdata/ocode_new", package = "gmoTree"),
  fsource = "init",
  output = "list",
  doc_info = FALSE,
  app_rm = "prisoner")

str(cbook, 1)
## List of 15
##  $ settings            :List of 10
##  $ bargaining          :List of 4
##  $ bertrand            :List of 4
##  $ common_value_auction:List of 4
##  $ cournot             :List of 4
##  $ dictator            :List of 4
##  $ guess_two_thirds    :List of 4
##  $ matching_pennies    :List of 4
##  $ payment_info        :List of 4
##  $ public_goods_simple :List of 3
##  $ survey              :List of 3
##  $ traveler_dilemma    :List of 4
##  $ trust               :List of 3
##  $ trust_simple        :List of 3
##  $ volunteer_dilemma   :List of 4

oTree data description

By setting the preamb = TRUE parameter, a predefined description of oTree data is inserted at the beginning of the codebook, explaining how experimental data is stored by oTree.

Introduction to oTree data picture

Customizing how the codebook is shown

Customizing the codebook titles and metadata

You can easily modify the titles, subtitles, dates, and author names in your codebooks using the following arguments:

codebook(
    path = "YouroTreePath",  # Can be either absolute or relative!
    title = "Codebook",
    subtitle = "Example Project",
    date = "today",
    params = list(author = c("Max Mustermann", "John Doe"))
)

Image showing the title section in a PDF output

Sorting the code

You can sort the apps in the codebook by specifying the sort argument. For example:

cbook <- codebook(
  path = system.file("extdata/ocode_new", package = "gmoTree"),
  fsource = "init",
  output = "list",
  doc_info = FALSE,
  app = c("cournot", "dictator"),
  sort = c("dictator", "cournot"))

str(cbook, 1)
## List of 3
##  $ settings:List of 10
##  $ dictator:List of 4
##  $ cournot :List of 4

Referring to settings

The gmoTree package can automatically replace references to variables in settings.py with the values in the settings.py file. To enable this, set settings_replace = "global".

cbook <- codebook(
  path = system.file("extdata/ocode_new", package = "gmoTree"),
  fsource = "init",
  output = "list",
  settings_replace = "global",
  app = "dictator")
## Variables without documentation, label, or verbose name:
## > $dictator$Player$gender
print(cbook$dictator$Constants$Variable)
## [1] 66

You can also choose to not replace the references to settings variables by using settings_replace = NULL.

cbook <- codebook(
  path = system.file("extdata/ocode_new", package = "gmoTree"),
  fsource = "init",
  output = "list",
  settings_replace = NULL,
  app = "dictator")
## Variables without documentation, label, or verbose name:
## > $dictator$Player$gender
print(cbook$dictator$Constants$Variable)
## [1] "settings.variable"

It is also possible to provide a list of variables to replace the settings variables with the argument user_settings. For example:

cbook <- codebook(
  path = system.file("extdata/ocode_new", package = "gmoTree"),
  fsource = "init",
  output = "list",
  settings_replace = "user",
  user_settings = list(variable = "someValue"),
  app = "dictator")
## Variables without documentation, label, or verbose name:
## > $dictator$Player$gender
print(cbook$dictator$Constants$Variable)
## [1] "someValue"

Saving your codebook

When you choose output = "file" or output = "both", your codebook will be saved as a professional document ready for publication or sharing.

Output formats

The argument output_format allows you to choose the format of the exported codebook file. Available formats are:

You can choose the format that best suits your needs, with all formats except PDF allowing easy post-generation editing.

When knitting to PDF, make sure to have LaTex installed! It is recommended to use output_format = pdf_document_simple. However, many non-Latin characters (e.g., Chinese characters) may not work with this format. In such cases, you can choose output_format = pdf_document, which is set to use xelatex.

Please note that using xelatex with gmoTree has some limitations: for example, gmoTree currently cannot handle long variable values that result in excessively long table cells. This can cause the PDF to display improperly in certain viewers, such as Nitro. To address this, carefully review your file. If necessary, consider using output_format = latex_document. Alternatively, open the file with a PDF viewer that can handle these cases, then save it there to ensure it displays correctly in other PDF viewers.

Output files

By default, gmoTree codebooks are saved in your working directory with the default file name codebook. You can modify this by specifying a custom file name and directory path.

For example, to save the codebook with a different name, use the following code:

codebook(
    path = "C:/Users/username/folder/oTree",  # Replace!
    output_file = "gmoTree_codebook")

You can also include the file extension, although it is optional:

codebook(
    path = "C:/Users/username/folder/oTree",  # Replace!
    output_file = "gmoTree_codebook.pdf")

You can save the file in a subfolder within the current directory by including the absolute or relative folder name in the output_file argument. For example:

codebook(
    path = "C:/Users/username/folder/oTree",  # Replace!
    output_file = "codebookfiles/gmoTree_codebook.pdf")

You can also specify absolute paths, either in the output_dir argument or directly within the file name specified in the output_file. For example:

codebook(
    path = "C:/Users/username/folder/oTree",  # Replace!
    output_dir  = "C:/Users/username/folder/codebooks",
    output_file = "mycodebook")

Summary

This vignette documents how to customize the codebook generated by the codebook() function, including options for content selection, formatting, and exporting to various formats. These tools enable the creation of comprehensive and tailored documentation for your experimental data.

References

Chen, D. L., Schonger, M., & Wickens, C. (2016). oTree—An open-source platform for laboratory, online, and field experiments. Journal of Behavioral and Experimental Finance, 9, 88–97. https://doi.org/10.1016/j.jbef.2015.12.001


  1. The oTree sample codes can be accessed when setting up an oTree project (Chen et al., 2016). All examples presented in this vignette are based on the official oTree sample code, with minor modifications to meet the needs of this package.↩︎