Conditionally change column value based on value in another column

Question

I have a dataset in R that is structured like so:

> head(top10_master, n = 5)
   sensitivity      PRODUCT NewRevenue NewQuantity NewProfit Change_Price Rev_Change Prof_Change Quant_Change Revenue Quantity  Profit
1:           1 PRODUCT33411    7067750    11809634   5790148      0.80000  1.2282115    1.173265    1.5591315 5754505  7574495 4935074
2:           1 PRODUCT35481    4484178     5721789   3278738      1.04978  1.0000000    1.000000    1.0000000 4484178  5721789 3278738
3:           1 PRODUCT26835    5239519     2239783   3950871      0.90000  1.1801825    1.125136    1.3884444 4439584  1613160 3511460
4:           1 PRODUCT33203    5872765     1139270   3125781      0.90000  1.3477930    1.161484    1.6487265 4357320   691000 2691195
5:           1 PRODUCT22426    4125707     1121494   3249458      0.80000  0.9590297    1.378040    0.4507613 4301960  2488000 2358029

I am trying to edit the Change_Price column to 1 (inside of a function) in the instances where the Rev_Change and Quant_Change columns == 1. I have tried every method I could find including using dplyr but for some reason none of these methods are working and the output is always the same. Please help.

I have tried too many methods to list them all but among them are all of the methods listed in this post. For some reason, none of these methods is working.

@akrun I suspect you may be right because it seems like traditional subsetting is not recognizing `Change_Rev = 1` etc... how would I fix that? — zsad512, Dec 21 '17 at 19:16

score 0 · Accepted Answer · answered Dec 22 '17 at 00:17

0

As akrun suggested, maybe there is a floating point accuracy issue. You could try using dplyr's near:

This is a safe way of comparing if two vectors of floating point numbers are (pairwise) equal. This is safer than using ==, because it has a built in tolerance.

This might work for you:

library(dplyr)
top10_master$Change_Price[which(near(top10_master$Rev_Change, 1) & near(top10_master$Quant_Change, 1))] <- 1

answered Dec 22 '17 at 00:17

Hallie Swan

2,714
1
15
23

the only problem with this solution is that it also seems to round the other values as well – zsad512 Dec 22 '17 at 19:47
@zsad512 I don't fully understand your issue. Can you post an example? – Hallie Swan Dec 22 '17 at 20:00

troh · Answer 2 · 2017-12-21T22:32:17.857

-1

top10_master$Change_Price[sapply(top10_master$Rev_Change, all.equal, 1) & sapply(top10_master$Quant_Change, all.equal, 1)] <- 1

Subset to the data you want to change and then assign.

edited Dec 21 '17 at 22:32

answered Dec 21 '17 at 18:58

troh

1,354
10
19

as mentioned in my post @troh I have tried this and it didnt work – zsad512 Dec 21 '17 at 19:17
@zsad512 I've updated my answer from the comment above. – troh Dec 21 '17 at 22:32
`Error in sapply(inData$Change_Rev, all.equal, 1) & sapply(inData$Change_Quant, : operations are possible only for numeric, logical or complex types` – zsad512 Dec 21 '17 at 23:08
I checked the input data tables as well before posting ^ and they are all listed as `numeric` so not really sure whats going on here... – zsad512 Dec 21 '17 at 23:14
I'm assuming you used: `is.numeric(inData$Change_Rev); is.numeric(inData$Change_Quant)` – troh Dec 22 '17 at 14:32

Conditionally change column value based on value in another column

2 Answers2