R Code examples

R code snippets for table manipulation

In R language dplyr and tidyr are two useful packages for handling and manipulating tables.

R code snippet to join two tables

library(dplyr) 

# Read csv tables 
tbl_subjects <- read.csv('dummy_table_OK.csv')
tbl_files <- read.csv('dummy_table_filenames.csv')

# Join by variable "Subject_ID" 
tbl_joined <- dplyr::full_join(x=tbl_subjects, 
                               y = tbl_files, 
                               by=join_by("Subject_ID"))

Table 1: Table with file information

Subject_ID	Files	Session
S001	filename1.ext	1
S002	NA	NA
S003	filename1.ext	1
S003	filename2.ext	1
S003	filename3.ext	2
S003	filename4.ext	2

Table 2: Combined table after adding subject information

Subject_ID	Scan_Date	Modality	Sex	Age_years	Weight_kg	Exp_Group	Scanner_Model	Notes	Files	Session
S001	01/01/2023	MRI	M	25	70.00	Control	Siemens	None	filename1.ext	1
S002	02/01/2023	fMRI	F	26	72.04	Exp	Siemens	Motion artifact	NA	NA
S003	03/01/2023	DTI	M	27	75.00	Exp	Philips	High noise	filename1.ext	1
S003	03/01/2023	DTI	M	27	75.00	Exp	Philips	High noise	filename2.ext	1
S003	03/01/2023	DTI	M	27	75.00	Exp	Philips	High noise	filename3.ext	2
S003	03/01/2023	DTI	M	27	75.00	Exp	Philips	High noise	filename4.ext	2

In more complex settings, it is possible to specify the expected relationship between two tables (e.g., if each row from table x is expected to have a match in table y, etc.) see mutate-joins.

R code to convert from wide to long format

library(tidyr) 
library(dplyr) 

# Make an example table
tbl_wide <- data.frame(
  Subject_ID = c("S001", "S002", "S003", "S004"),
  Test_A = c(5.4, 6.1, 7.3, 4.8),
  Test_B = c(3.2, 5.8, 6.5, 4.1)
)

# Convert to long format
tbl_long <- tbl_wide %>% 
  pivot_longer(cols = c(Test_A, Test_B), names_to = "Test",
               values_to = "Value")

R code to convert from long to wide format

library(tidyr)
library(dplyr) 

tbl_wide_again <- tbl_long %>% 
                  pivot_wider(names_from = Test, values_from = Value) 

R code snippet to create JSON files from a table

In this example each row of a table with metadata is transformed into a JSON file. The R package jsonlite offers several encoding formatting options.

• A dummy metadata table

id	img_location	sex	condition	treatment
subject_04	dummyfolder/subject_04_imgfile.tiff	female	A	control
subject_93	dummyfolder/subject_93_imgfile.tiff	female	B	control
subject_12	dummyfolder/subject_12_imgfile.tiff	female	B	treat_3
subject_65	dummyfolder/subject_65_imgfile.tiff	male	B	control
subject_98	dummyfolder/subject_98_imgfile.tiff	female	A	treat_1
subject_71	dummyfolder/subject_71_imgfile.tiff	male	A	treat_3
subject_24	dummyfolder/subject_24_imgfile.tiff	female	B	control

• The rows converted to JSON

library(jsonlite)

# Write each row to a JSON object
row_json <- list()
for (i in 1:nrow(metadata_table)) {
  row_json[[i]] <- toJSON(metadata_table[i, ], 
                          pretty = TRUE,
                          auto_unbox = TRUE)
} 

The following line can be added in the loop to save the row as a JSON file with the same name as the images listed in the table

   img_file_name <- basename(metadata_table$img_location[i])  
   json_file_name <- sub("\\.tiff$", ".json", img_file_name)  
  write(row_json[[i]], file = json_file_name)