Updating old column entries from new data frame - r

Question

I am working on a problem. Here is an idea of what the original 60k row data frame looks like.
dataOne <- data.frame(
marketVal = c(NA, 543534, NA, 115435, NA),
bathrooms = c(3,3,2,3,5),
garageSqFt = c(400, 385, 454, 534, 210),
totalSqFT = c(NA, NA, 1231, 2232, 4564),
units = c(1, 1, 1, 1, 1),
subDivId = c("112", "111", "111", "111", "112"),
ID = c(4,56,67,94,130) )
Some of the NA's for market value have been retrieved and stored in a new
data frame that looks like so:
dataTwo <- data.frame(
marketVal = c(123123,234234),
IDTwo = c(4,67) )
str(dataTwo)
dataOne$marketVal <- dataTwo$marketVal[match(dataTwo$ID, dataOne$ID)]
comparing ID's from both data frames I am attempting to replace the NA's in the first data frame with the market values in the second data frame. I've tried the match function as follows:
dataOne$marketValue <- dataTwo$marketValue[match(dataOne$ID, dataTwo$ID)]
but recieve an error "replacement has 2 rows, data has 5 calls". I fugured the fact these two data frames not being the same size wouldn't matter as we are only comparing the ID's found in either. How can I accomplish this efficiently considering around 4500 NA's need to be updated?

Answer 1

Your method isn't working because it is producing a vector with 5 values: 1 NA 2 NA NA which is longer than your dataTwo dataframe. Drop the NA values and your method would work.
This is how I would do it:
rowMatch <- which(dataOne$ID %in% dataTwo$ID)
dataOne[rowMatch, ]$marketVal <- dataTwo$marketVal
(Please note your ID variables were actually IDOne and IDTwo respectively in the example you'd provided.)

Answer 2

You can use merge
require(tidyverse)
new <- merge(dataOne, dataTwo, by.x = 'ID', by.y = 'IDTwo', all.x = T)
new$marketVal <- new %$% coalesce(marketVal.x, marketVal.y)

Answer 3

We could use safe_left_join from my package safejoin, and "patch"
the matches from the rhs into the lhs when columns conflict.
# devtools::install_github("moodymudskipper/safejoin")
library(safejoin)
library(dplyr)
dataOne <- data.frame(
marketVal = c(NA, 543534, NA, 115435, NA),
bathrooms = c(3,3,2,3,5),
garageSqFt = c(400, 385, 454, 534, 210),
totalSqFT = c(NA, NA, 1231, 2232, 4564),
units = c(1, 1, 1, 1, 1),
subDivId = c("112", "111", "111", "111", "112"),
ID = c(4,56,67,94,130) )
dataTwo <- data.frame(
marketVal = c(123123,234234),
IDTwo = c(4,67) )
safe_left_join(dataOne, dataTwo, by=c(ID= "IDTwo"), conflict = "patch")
# marketVal bathrooms garageSqFt totalSqFT units subDivId ID
# 1 123123 3 400 NA 1 112 4
# 2 543534 3 385 NA 1 111 56
# 3 234234 2 454 1231 1 111 67
# 4 115435 3 534 2232 1 111 94
# 5 NA 5 210 4564 1 112 130
or for the same effect in this case we can use dplyr::coalesce
library(dplyr)
safe_left_join(dataOne, dataTwo, by=c(ID= "IDTwo"), conflict = coalesce)