irwpkg

irwpkg is an R package designed to facilitate easy access to datasets from the Item Response Warehouse (IRW), offering functions to download, explore, and analyze these datasets.

Installation

You can install the development version of irwpkg from GitHub with:

# install.packages("remotes")
remotes::install_github("hansorlee/irwpkg")

Usage

To use irwpkg, load the package using:

library(irwpkg)

Authentication

The IRW datasets are hosted on Redivis, so to access these datasets through irwpkg, you’ll need to authenticate with your Redivis account. This connection is managed through the Redivis R client.

Redivis Authentication Setup

When you first use a function in irwpkg that connects to Redivis (e.g. see list_available_datasets() below), a browser window will pop up prompting you to sign into Redivis.
After signing in, you’ll be prompted to allow access for the Redivis R Client. Click Allow.
Once you’ve granted access, close the browser window. You’ll see the message “Authentication was successful” in the R console, and you can then re-run your irwpkg function.

Note: Authentication typically only needs to be done once per session.
For more details, refer to Redivis R Client documentation.

View list of datasets

The list_available_datasets() function provides a list of datasets in the IRW database, showing key metadata (dataset name, row count, and variable count).

# List available datasets
list_datasets = list_available_datasets()
dim(list_datasets)
## [1] 558   3
head(list_datasets)
##                                 name numRows variableCount
## 1                        lessR_Mach4    7020             3
## 2                  movac_pakpour2022   72096             3
## 3 SABFI2_Gallardo_Pujol_2018_IntHapp    3771             3
## 4                        science_ltm    2744             4
## 5                    gilbert_meta_17 1840581            11
## 6                    pks_probability   12096             4

view DB/table metadata

db_info = get_database_metadata()
## IRW Database Metadata:
## --------------------------------------------------
## Version:                   v11.19 
## Table Count:               558 
## Created At:                2023-09-25 19:19:15 
## Last Updated At:           2024-11-14 16:31:26 
## Total Data Size (GB):      145.1 GB
## Active Data Size (GB):     145.01 GB
## DOI:                       https://doi.org/10.57761/4zaa-w743 
## Dataset URL:               https://redivis.com/datasets/as2e-cv7jb41fd?v=11.19 
## Documentation:             https://datapages.github.io/irw/ 
## Methodology:               Tables have been harmonized as per details given [here](<https://datapages.github.io/irw/standard.html>).
## 
## Usage Information:         Please find information about data licenses and citation info [here](<https://datapages.github.io/irw/docs.html>).
## 
##  
## --------------------------------------------------

table_info = get_table_metadata("lessR_Mach4")
## Table Metadata for: lessR_Mach4 
## --------------------------------------------------
## Name:                      lessR_Mach4 
## Created At:                2024-11-14 16:25:32 
## Last Updated At:           2024-11-14 16:31:26 
## Number of Rows:            7020 
## Data Size (KB):            141.86 KB
## Variable Count:            3 
## Is Sample:                 No 
## DOI:                       https://doi.org/10.57761/4zaa-w743 
## Table URL:                 https://redivis.com/datasets/as2e-cv7jb41fd/tables/018s-9p87q43mh?v=11.19 
## Container URL:             https://redivis.com/datasets/as2e-cv7jb41fd?v=11.19 
## --------------------------------------------------

filter tables by criteria

# get tables with a column named "rater"
filter_tables(required_columns="rater")
##  [1] "swmd_mokken"                    "dumas_Organisciak_2022"        
##  [3] "fractals_rating"                "mpsycho_lakes"                 
##  [5] "ptam1_immer"                    "autonomysupport_mokken"        
##  [7] "wine_luckett2021"               "spelling2pronounce_edwards2023"
##  [9] "immer12_immer"                  "famous_melodies"

# get tables with at least 10000 rows and a column named "rater"
filter_tables(n_rows = 10000, required_columns="rater")
## [1] "dumas_Organisciak_2022"         "fractals_rating"               
## [3] "mpsycho_lakes"                  "ptam1_immer"                   
## [5] "spelling2pronounce_edwards2023" "famous_melodies"

# get tables with columns "rater" and "rt"
filter_tables(required_columns=c("rater", "rt"))
## [1] "famous_melodies"

Fetch data

Fetch a single dataset

Once you have identified a dataset you want, you can use fetch_data() to load it as a data frame in R. For example, to fetch the swmd_mokken dataset, you can use:

# Fetch a dataset by name
swmd_mokken <- fetch_data(name="swmd_mokken")
dim(swmd_mokken)
## [1] 4557    4
head(swmd_mokken)
##   resp  item rater      id
## 1    1 Item2    12 1000303
## 2    1 Item1    13 1000303
## 3    1 Item3    17 1001302
## 4    1 Item1    22 1001302
## 5    1 Item2    25 1001302
## 6    1 Item3    25 1001302

Fetch multiple datasets

You can also fetch multiple datasets at once.

matching_tables <- filter_tables(n_rows = 50000, required_columns = c("rater"))
print(matching_tables)
## [1] "fractals_rating"                "spelling2pronounce_edwards2023"

datasets <- fetch_data(c("fractals_rating", "spelling2pronounce_edwards2023"))
print(names(datasets)) # datasets is a named list
## [1] "fractals_rating"                "spelling2pronounce_edwards2023"

You can also use the output of filter_tables() directly to fetch_data().

datasets <- fetch_data(matching_tables)
print(names(datasets))
## [1] "fractals_rating"                "spelling2pronounce_edwards2023"

fractals_rating <- datasets[["fractals_rating"]]
head(fractals_rating)
##            id            item rater resp
## 1 Fractal_001      liveliness  1012    1
## 2 Fractal_001 verbalizability  1019    1
## 3 Fractal_001  favourableness  1007    1
## 4 Fractal_001     familiarity  1009    1
## 5 Fractal_001     familiarity  1019    1
## 6 Fractal_001 verbalizability  1007    1

Download data

download_data("swmd_mokken", path = "mydata.csv")

hansorlee / irwpkg

readme