Replace some rows in data frame with other new data frame using dplyr in R-CodePudding

I have data that call mydata. mydata contains full september data from Sales Alabama Store. Here it is:

# My Data
library(lubridate)
gen_seq <- seq(ymd_h("2021-09-01-00"), ymd_h("2021-09-30-23"), by = "hours")
datex <- date(gen_seq)
hourx <- hour(gen_seq)
store <- "Alabama"
sales <- round(runif(length(datex), 10, 50), 0)
mydata <- data.frame(datex, hourx, store, sales)

And now i have a new revision from date "2021-09-15" & "2021-09-16", It is the revision:

# New My Data
library(lubridate)
gen_seq <- seq(ymd_h("2021-09-15-00"), ymd_h("2021-09-16-23"), by = "hours")
datex <- date(gen_seq)
hourx <- hour(gen_seq)
store <- "Alabama"
sales <- round(runif(length(datex), 10, 50), 0)
newmydata <- data.frame(datex, hourx, store, sales)

How do i replace it? Thanks

CodePudding user response：

You can join the two datasets and update the values with coalesce.

library(dplyr)

left_join(mydata, newmydata,  by = c("datex", "hourx", "store")) %>%
  mutate(sales = coalesce(sales.y, sales.x)) %>%
  select(-sales.x, -sales.y)

In base R,

merge(mydata, newmydata, by= c("datex", "hourx", "store"), all.x = TRUE) |>
  transform(sales = ifelse(is.na(sales.y), sales.x, sales.y))