Manual reference retrieval

If you are here, this is because you want to manually retrieve a list of references, outside of the automated task that you might have set up after reading LitFetchR (Tutorial).

Setup

From your individual review R directory, load LitFetchR:

# Load LitFetchR
library(LitFetchR)

Manual reference retrieval

Make sure to have saved your search string using create_save_search() as the following function needs to access the file search_list.txt, see LitFetchR (Tutorial).

The optional deduplication (dedup = TRUE) needs the package ASySD, which is not on CRAN. Install it with remotes::install_github("camaradesuk/ASySD"), or use dedup = FALSE.

# Load LitFetchR
library(LitFetchR)

# Retrieve the list of references corresponding to your saved search string(s)
#We recommend choosing the R project directory as the directory in the function. It has to be the directory where "search_list.txt" and "history_search.xlsx" have been saved.

# Example of what you should see (a later run, the history files already exist):
# > manual_fetch(wos = TRUE,
#                scp = TRUE,
#                pmd = TRUE,
#                directory = "C:/path-to-project-directory",
#                dedup = TRUE,
#                open_file = FALSE)
# 154 total results found on Web of Science for: fish AND "vibrio harveyi" AND diagnostic
# Finished batch 1 of 2
# Finished batch 2 of 2
# Updating existing history_id.xlsx.
# 149 new records found among 154 total results.
# 24 total results found on Scopus for: fish AND "vibrio harveyi" AND diagnostic
# Finished batch 1 of 1
# Updating existing history_id.xlsx.
# 22 new records found among 24 total results.
# 110 total results found on PubMed for: fish AND "vibrio harveyi" AND diagnostic
# Finished batch 1 of 1
# Updating existing history_id.xlsx.
# 104 new records found among 110 total results.
# Finished batch 1 of 1 (104 / 104 records retrieved)
# Warning: The following columns are missing: record_id
# formatting data...
# identifying potential duplicates...
# identified duplicates!
# 275 citations loaded...
# 27 duplicate citations removed...
# 248 unique citations remaining!
# Deduplication script has been executed, concatenated deduplicated references had been exported.
# Warning message:
# In add_missing_cols(raw_citations) :
#   Search contains missing values for the record_id column. A record_id will be created using row numbers
#
# Run log saved in C:/path-to-project-directory/run_log.csv

After running manual_fetch(), new unique history_dedup_ and citationsCSV_ files are created, history_id is updated and one row per saved search and platform is added to run_log.csv. When no reference is new, the function reports it (No new records: no output file written.) and no output file is written. See bottom of the page LitFetchR (Tutorial) for more information about those files.

Output format

By default the references are written to a CSV file (citationsCSV_YYYY_MM_HHMMSS.csv). Some tools do not accept CSV, for example Covidence, which imports RIS files. Use output_format = "ris" to get a RIS file (citationsRIS_YYYY_MM_HHMMSS.ris) instead:

manual_fetch(wos = TRUE,
             scp = TRUE,
             pmd = TRUE,
             directory = "C:/path-to-project-directory",
             dedup = TRUE,
             output_format = "ris")

The RIS file has one record per reference, with one author line per author. The platforms give the authors as surnames, so initials are not included. The same argument exists in auto_LitFetchR_setup(). A scheduled task keeps the format it was created with: to change it, remove the task and create it again.

If something goes wrong