How to delete rows given conditions in R?

Question

So I know how to delete rows based off one column condition. How do I do it for multiple conditions?

Date Frame = "olddata"

Letter Number Color
X 1 Blue
Y 3 Red
Z 5 Yellow
D 1 Red
F 1 Yellow
X 2 Green

For example, to delete rows with the Letter X in it, and rows with color green, you can do:

newdata <- subset(olddata, olddata$Letter != "X" & olddata$Color != "Green")

Now how do I delete Rows that have the letter F and is Yellow, i.e. both criteria must be met.

Thanks!

Combine your criteria and then negate it - `olddata[!(olddata$Letter == "F" & olddata$Color == "Yellow"),]` — thelatemail, Feb 26 '18 at 21:48
already saw that question and the formula didn't work for me: — user6472523, Feb 26 '18 at 21:53
Are you sure, because the answer there is identical in logic to my comment above, and I can confirm it works with your current data. — thelatemail, Feb 26 '18 at 21:55
`subset` will allow multiple criteria for sure - `subset(mtcars, cyl == 4 & mpg > 30)` for a reproducible example. — thelatemail, Feb 26 '18 at 21:59
Look carefully. The difference between the answer below and the one given by @thelatemail is in parentheses. thelatemail correctly negates the entire condition, while the answer below does it to the first term only. — Artem Sokolov, Feb 26 '18 at 22:03
in this: subset(mtcars, cyl == 4 & mpg > 30) I'm pretty sure that removes any row that has either cyl = 4 and or mpg >30. It does not remove just the rows that are BOTH cyl = 4 and mpg >30. At least that what happens when I used that formula — user6472523, Feb 28 '18 at 15:11
@user6472523 - *"I'm pretty sure that removes any row that has either cyl = 4 and or mpg >30"* - actually it *keeps* any row that has `cyl==4` **and** `mpg>30`. You'd use `subset(mtcars, !(mtcars$cyl == 4 & mtcars$mpg > 30))` if you wanted to remove the rows. — thelatemail, Mar 04 '18 at 23:58

score 0 · Answer 1 · edited Feb 26 '18 at 22:30

0

newdata <- olddata[!(olddata$Letter == "F" & olddata$Color == "Yellow"),]

edited Feb 26 '18 at 22:30

thelatemail

answered Feb 26 '18 at 21:43

AidanGawronski

That would remove both rows 3 and 6. It's just like my old formula. I want to remove just row 6. – user6472523 Feb 26 '18 at 21:44
If you're going to use `subset`, no need for `olddata$` again each time. – thelatemail Feb 26 '18 at 21:47
@thelatemail, that's good to know. I prefer base R and I assumed subset basically functioned the same way, changing my answer to base R. – AidanGawronski Feb 26 '18 at 21:54
No problems. `subset` is base R by the way, just a 'non-standard evaluation' function that R provides. Also, I think you need to negate your whole selection - at the moment you're keeping "F"/"Yellow" rows rather than deleting them. – thelatemail Feb 26 '18 at 21:57
lol... yes ... yes I am. – AidanGawronski Feb 26 '18 at 21:59
Just to continue to be a nuisance - `!olddata$Letter == "F" & olddata$Color == "Yellow"` is not the same as `!(olddata$Letter == "F" & olddata$Color == "Yellow")` - check each result in the console (I didn't downvote btw). – thelatemail Feb 26 '18 at 22:02
I feel like I'm in a slap stick routine. lol. – AidanGawronski Feb 26 '18 at 22:06

score -2 · Answer 2 · edited Feb 26 '18 at 22:37

-2

Use:

library(dplyr)

olddata %>%
filter(olddata, Letter!="F" & Color!="Yellow")

edited Feb 26 '18 at 22:37

Daniel V

answered Feb 26 '18 at 21:47

Hello.World

This doesn't give the right result. – thelatemail Feb 26 '18 at 21:49
Do you mean `require(dplyr); df %>% filter(!(Letter == "F" & Color == "Yellow"));`? – Maurits Evers Feb 26 '18 at 22:01

2 Answers2