gmoTree Codebooks

Patricia F. Zauchner

2023-09-21 (updated: 2024-09-30)

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 you to the key features and functionalities, including how to adjust titles, work with oTree data, and customize output formats for your codebooks.

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 code reads the contents of your oTree file, saves a file containing your codebook in your working directory, and returns a list named cbook, which contains all the information in a list format.

cbook <- codebook(system.file(path = "extdata/ocode_new", 
                              package = "gmoTree"))

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

Below is an example showing 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

If the default codebook design does not meet your requirements, you can either reference the previously created cbook list or bypass file creation by setting output = "list". This will return all the information only in list format, which you can then use in R Markdown to design your own custom codebook. Below is an example of a section of the list output for the cournot app. (The argument doc_info = FALSE is used to omit messages about missing variable documentation.)

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

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      : chr "[]"
##   ..$ SESSION_FIELDS          : chr "[]"
##   ..$ 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: chr "TRUE"
##   .. .. ..$ field : chr "CurrencyField"
##   .. ..$ total_units:List of 3
##   .. .. ..$ noargs: chr "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 3

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 3
##  $ 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 releative!
    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 3
##  $ 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.

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

print(cbook$dictator$Constants$Variable)
## [1] 66

However, you can also choose to not replace them by using settings_replace = NULL:

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

print(cbook$dictator$Constants$Variable)
## [1] "settings.variable"

It is also possible to provide a list of variables that should replace the settings variables:

cbook <- codebook(
  path = system.file("extdata/ocode_new", package = "gmoTree"),
  fsource = "init",
  output = "list",
  settings_replace = "user",
  user_settings = list(variable = "someValue"),
  app = "dictator")

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.

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 offering easy post-generation editing.

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/Nextcloud/oTree",  # Replace!
    output_file = "gmoTree_codebook")

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

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

However, 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/Nextcloud/oTree",  # Replace!
    output_dir  = "C:/Users/username/Nextcloud/codebooks",
    output_file = "mycodebook")

Summary

This vignette covers how to customize the codebook generated by the codebook() function, including options for selecting content, formatting its presentation, and exporting it in 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.↩︎