Reputation: 191
the dataset is something like this:
daily.sample
day1 day2 day3 day4 day5 day6 day7 day8 day9 day10 day11 day12 day13 day14
132304 919 420 1021 331 1012 606 256 58 302 87 859 1140 603 1472
121945 419 1048 541 336 1086 796 95 578 1419 985 752 570 865 242
130055 3036 1473 3043 3291 4869 2902 5306 4480 4820 254 2652 2529 3346 995
I want to select rows that satisfy the following conditions: it's day1 value is close to it's day8 value it's day2 value is close to it's day9 value it's day3 value is close to it's day10 value ...
by close i mean the difference is within a range, or something like "0.9< day1/day8 <1.1"
Upvotes: 0
Views: 2358
Reputation: 7113
You can simply substract the right part of your data.frame
from the left part by:
distance <- abs(daily.sample[ , 1:7, drop = FALSE ] - daily.sample[ , 8:14, drop =FALSE ])
The ratio is build accordingly:
ratio <- daily.sample[ , 1:7, drop = FALSE ] / daily.sample[ , 8:14, drop =FALSE ]
And then you can simply subset your data depending on the day for which you want to check.
daily.sample[ distance[ ,"day1" ] < 1000, ]
day1 day2 day3 day4 day5 day6 day7 day8 day9 day10 day11 day12 day13 day14
132304 919 420 1021 331 1012 606 256 58 302 87 859 1140 603 1472
121945 419 1048 541 336 1086 796 95 578 1419 985 752 570 865 242
Or for the ratio within a range:
daily.sample[ ratio[ , "day1"] > 0.7 & ratio[ , "day1"] < (10/7), ]
day1 day2 day3 day4 day5 day6 day7 day8 day9 day10 day11 day12 day13 day14
121945 419 1048 541 336 1086 796 95 578 1419 985 752 570 865 242
If for the latter case the range is always (min.ratio, 1/min.ratio) you could calculate the ratio as ratio[ ratio < 1 ] <- 1 / ratio[ ratio < 1 ]
and than just subset by:
daily.sample[ ratio[ , "day1"] < (10/7), ].
Upvotes: 1