Summing over columns with if statement in R - r

Question

I have a data that looks like this.
bankname bankid year totass corresbankname1 corresbankloc1 corresdepoin1 corresbankname2 corresbankloc2 corresdepoin2 corresbankname3 corresbankloc3 corresdepoin3
BankA 1 1881 244789 First Bank New York 7250.32 Third National Bank Philadelphia 20218.2 Commercial Bank Philadelphia 29513.4
BankB 2 1881 195755 National Bank Pittsburgh 10243.6 Union Trust Company New York 1851.51 NA NA NA
Bankc 3 1881 107736 Mechanics' Bank New York 13357.8 Wyoming Bank Wilkes-Barre 17761.2 NA NA NA
BankD 4 1881 170600 Commonwealth Bank Philadelphia 3.35 Seventh National Bank Philadelphia 2 NA NA NA
BankE 5 1881 320000 National Bank New York 351266 Mechanics' Bank New York 314012 National Park Bank New York 206580
This can be replicated by
bankname <- c("The Anchor Savings Bank of Pittsburgh","The Arsenal Bank","The Ashley Savings Bank","The Bank of America of Philadelphia","The Bank of Pittsburgh")
bankid <- c( 1, 2, 3, 4, 5)
year<- c( 1881, 1881, 1881, 1881, 1881)
totass <- c(244789, 195755, 107736, 170600, 32000000)
corresbankname1 <- c("First National Bank","National Bank of Commerce","Mechanics' National Bank","Commonwealth National Bank","National Bank of Commerce")
corresbankloc1 <-c("Philadelphia","Pittsburgh","New York","Philadelphia","New York")
corresdepoin1<-c(7250.32,10243.6,13357.8,3.35,351266)
corresbankname2 <- c("Third National Bank","Union Trust Company","Wyoming National Bank","Seventh National Bank","Mechanics' National Bank")
corresbankloc2<-c("New York","New York","Wilkes-Barre","Philadelphia","New York")
corresdepoin2<-c(20218.2,1851.51,17761.2,2,314012)
corresbankname3<-c("Commercial National Bank",NA,NA,NA,"National Park Bank")
corresbankloc3<-c("Philadelphia",NA,NA,NA,"New York")
corresdepoin3<-c(29513.4,NA,NA,NA,206580)
bankdata<-data.frame(bankname, bankid,year,totass,corresbankname1,corresbankloc1,corresdepoin1,corresbankname2,corresbankloc2,corresdepoin2,corresbankname3,corresbankloc3,corresdepoin3)
This dataset shows the amount(corresdepoin) each bank invested in other banks (corresbankname) and its location (corresbankloc). I have 43 corresbankname, corresbankloc and corresdepoin variables.
Since these banks invest in multiple banks in the same city, I would like to know the total amount of investments in each city. Hence, I would like to generate a new column variable called "total_New York" and sum amount indicated by corresdepoin if correspobankloc is New York. How can I loop over 43 variables?
For instance, BankE has $351266 (corresdepoin1) at National Bank(corresbankname1) in New York(corresbankloc1), $314012 at Mechenics' Bank in New York, and $206580 at National Park Bank in New York. I want a new column called "total deposits in New York" showing that the total amount of investments in banks located in New York is $871858. Hence, what I want is a conditional statement that loops over columns (corresponbankloc) and checks if this is New York or others, and then sum associated values in "corresdepoin" to get aggregated amount of total investment in that city for each "bankname."
Also, in stata, if I want to do this for multiple cities, I would generate
local cities "New York" "Philadelphia" "Pittsburgh"
and loop over them. Is there a similar function in R?
Thank you in advance.

Answer 1

Another option would be to reshape the dataset. Using dplyr. You may create a function to output a subset of cities or for the whole city in the data. I don't know if this is efficient.
library(dplyr)
fun1 <- function(data, city, byloc = TRUE, allcity = TRUE) {
data1 <- reshape(data, idvar = "bankname", varying = list(grep("corresdepoin",
colnames(data)), grep("corresbankloc", colnames(data))), timevar = "Bankloc",
direction = "long", v.names = c("corresdepoin", "corresbankloc"))
data1 <- data1[!is.na(data1$corresbankloc), ]
row.names(data1) <- 1:nrow(data1)
funlocorNot <- function(data, city, grouploc = TRUE) {
dataF <- data %>%
filter(corresbankloc %in% city)
if (grouploc) {
dataF1 <- dataF %>%
group_by(bankname, corresbankloc, Bankloc) %>%
summarise(Total = sum(corresdepoin, na.rm = TRUE))
}
else {
dataF1 <- dataF %>%
group_by(bankname, corresbankloc) %>%
summarise(Total = sum(corresdepoin, na.rm = TRUE))
}
dataF1[, 2] <- paste("Totalbylocation", dataF1[, 2], sep = "_")
dataF1
}
funallCity <- function(data, grouploc = TRUE) {
if (grouploc) {
dataF1 <- data %>%
group_by(bankname, corresbankloc, Bankloc) %>%
summarise(Total = sum(corresdepoin, na.rm = TRUE))
}
else {
dataF1 <- data %>%
group_by(bankname, corresbankloc) %>%
summarise(Total = sum(corresdepoin, na.rm = TRUE))
}
dataF1[, 2] <- paste("Totalbylocation", dataF1[, 2], sep = "_")
dataF1
}
if (!allcity) {
if (byloc) {
funlocorNot(data1, city, TRUE)
}
else {
funlocorNot(data1, city, FALSE)
}
}
else {
if (byloc) {
funallCity(data1, TRUE)
}
else {
funallCity(data1, FALSE)
}
}
}
as.data.frame(fun1(bankdata, "New York", byloc=TRUE, allcity=FALSE))
# bankname corresbankloc Bankloc
#1 The Anchor Savings Bank of Pittsburgh Totalbylocation_New York 2
#2 The Arsenal Bank Totalbylocation_New York 2
#3 The Ashley Savings Bank Totalbylocation_New York 1
#4 The Bank of Pittsburgh Totalbylocation_New York 1
#5 The Bank of Pittsburgh Totalbylocation_New York 2
#6 The Bank of Pittsburgh Totalbylocation_New York 3
# Total
#1 20218.20
#2 1851.51
#3 13357.80
#4 351266.00
#5 314012.00
#6 206580.00
as.data.frame(fun1(bankdata, "New York", byloc=FALSE, allcity=FALSE))
# bankname corresbankloc Total
#1 The Anchor Savings Bank of Pittsburgh Totalbylocation_New York 20218.20
#2 The Arsenal Bank Totalbylocation_New York 1851.51
#3 The Ashley Savings Bank Totalbylocation_New York 13357.80
#4 The Bank of Pittsburgh Totalbylocation_New York 871858.00
as.data.frame(fun1(bankdata, c("New York", "Pittsburgh"), byloc=FALSE, allcity=FALSE))
as.data.frame(fun1(bankdata, byloc=TRUE, allcity=TRUE))