I would like to take a random sample of rows from a data.frame, apply a function to the subset, then take a sample from the remaining rows, apply the function to the new subset (with different parameters), and so on.
A simple example would be if 5% of a population dies each month, in month 2 I need the population minus those ones who died in time month 1.
I have put together a very verbose method of doing this involving where I save the IDs from the sampled rows, then subset them out from the data for the second period, etc.
library(data.table)
dt <- data.table(Number=1:100, ID=paste0("A", 1:100))
first<-dt[sample(nrow(dt), nrow(dt)*.05)]$ID
mean(dt[ID %in% first]$Number)
second<-dt[!(ID %in% first)][sample(nrow(dt[!(ID %in% first)]),
nrow(dt[!(ID %in% first)])*.05)]$ID
mean(dt[ID %in% c(first,second)]$Number)
dt[!(ID %in% first)][!(ID %in% second)] #...
Obviously, this is not sustainable past a couple periods. What is the better way to do this? I imagine this is a standard method but couldn't think what to look for specifically. Thanks for any and all input.
This shows how to "grow" a vector of items that have been sampled at a 5% per interval time course:
removed <- numeric(0)
for ( i in 1:10){
removed <- c(removed, sample( (1:100)[!(1:100) %in% removed], # items out so far
(100-length(removed))*.05)) # 5% of remainder
cat(c(removed, "\n")) # print to console with each iteration.
}
54 1 76 96 93
54 1 76 96 93 81 16 13 79
54 1 76 96 93 81 16 13 79 80 74 30 29
54 1 76 96 93 81 16 13 79 80 74 30 29 52 33 86 19
54 1 76 96 93 81 16 13 79 80 74 30 29 52 33 86 19 34 32 41 62
54 1 76 96 93 81 16 13 79 80 74 30 29 52 33 86 19 34 32 41 62 5 70 8
54 1 76 96 93 81 16 13 79 80 74 30 29 52 33 86 19 34 32 41 62 5 70 8 66 82 50
54 1 76 96 93 81 16 13 79 80 74 30 29 52 33 86 19 34 32 41 62 5 70 8 66 82 50 6 91 99
54 1 76 96 93 81 16 13 79 80 74 30 29 52 33 86 19 34 32 41 62 5 70 8 66 82 50 6 91 99 46 27 51
54 1 76 96 93 81 16 13 79 80 74 30 29 52 33 86 19 34 32 41 62 5 70 8 66 82 50 6 91 99 46 27 51 22 23 20
Notice that the actual number of items added to the list of "removals" will be decreasing.
Related
I would like to create new variables by dividing all pre-existing variables by each other
e.g.
X1/X1, X1/X2, X1/X3, X1/X4, X1/X5, X1/X6, X1/X7, X1/X8, X1/X9, X1/X10,
X2/X1, X2/X2, X2/X3, X2/X4, X2/X5, X2/X6, X2/X7, X2/X8, X2/X9, X2/X10,
X3/X1, X3/X2 ...
I started by trying to do each individually, as below, but I need to replicate this with multiple variable names so an automation (I assume a function/lapply) would be ideal.
ds$rom_3_5m <- (ds$roll_open_mean_3m/ds$roll_open_mean_5m)
ds$rom_3_10m <- (ds$roll_open_mean_3m/ds$roll_open_mean_10m)
ds$rom_3_15m <- (ds$roll_open_mean_3m/ds$roll_open_mean_15m)
ds$rom_3_30m <- (ds$roll_open_mean_3m/ds$roll_open_mean_30m)
ds$rom_3_60m <- (ds$roll_open_mean_3m/ds$roll_open_mean_60m)
ds$rom_3_120m <- (ds$roll_open_mean_3m/ds$roll_open_mean_120m)
ds$rom_3_240m <- (ds$roll_open_mean_3m/ds$roll_open_mean_240m)
ds$rom_3_480m <- (ds$roll_open_mean_3m/ds$roll_open_mean_480m)
ds$rom_3_960m <- (ds$roll_open_mean_3m/ds$roll_open_mean_960m)
ds$rom_3_1920m <- (ds$roll_open_mean_3m/ds$roll_open_mean_1920m)
ds$rom_3_3840m <- (ds$roll_open_mean_3m/ds$roll_open_mean_3840m)
ds$rom_3_7680m <- (ds$roll_open_mean_3m/ds$roll_open_mean_7680m)
ds$rom_3_15360m <- (ds$roll_open_mean_3m/ds$roll_open_mean_15360m)
ds$rom_3_30720m <- (ds$roll_open_mean_3m/ds$roll_open_mean_30720m)
ds$rom_3_61440m <- (ds$roll_open_mean_3m/ds$roll_open_mean_61440m)
ds$rom_3_122880m <- (ds$roll_open_mean_3m/ds$roll_open_mean_122880m)
ds$rom_3_245760m <- (ds$roll_open_mean_3m/ds$roll_open_mean_245760m)
ds$rom_3_491520m <- (ds$roll_open_mean_3m/ds$roll_open_mean_491520m)
#5m
ds$rom_5_3m <- (ds$roll_open_mean_5m/ds$roll_open_mean_3m)
ds$rom_5_10m <- (ds$roll_open_mean_5m/ds$roll_open_mean_10m)
ds$rom_5_15m <- (ds$roll_open_mean_5m/ds$roll_open_mean_15m)
ds$rom_5_30m <- (ds$roll_open_mean_5m/ds$roll_open_mean_30m)
ds$rom_5_60m <- (ds$roll_open_mean_5m/ds$roll_open_mean_60m)
ds$rom_5_120m <- (ds$roll_open_mean_5m/ds$roll_open_mean_120m)
ds$rom_5_240m <- (ds$roll_open_mean_5m/ds$roll_open_mean_240m)
ds$rom_5_480m <- (ds$roll_open_mean_5m/ds$roll_open_mean_480m)
ds$rom_5_960m <- (ds$roll_open_mean_5m/ds$roll_open_mean_960m)
ds$rom_5_1920m <- (ds$roll_open_mean_5m/ds$roll_open_mean_1920m)
ds$rom_5_3840m <- (ds$roll_open_mean_5m/ds$roll_open_mean_3840m)
ds$rom_5_7680m <- (ds$roll_open_mean_5m/ds$roll_open_mean_7680m)
ds$rom_5_15360m <- (ds$roll_open_mean_5m/ds$roll_open_mean_15360m)
ds$rom_5_30720m <- (ds$roll_open_mean_5m/ds$roll_open_mean_30720m)
ds$rom_5_61440m <- (ds$roll_open_mean_5m/ds$roll_open_mean_61440m)
ds$rom_5_122880m <- (ds$roll_open_mean_5m/ds$roll_open_mean_122880m)
ds$rom_5_245760m <- (ds$roll_open_mean_5m/ds$roll_open_mean_245760m)
ds$rom_5_491520m <- (ds$roll_open_mean_5m/ds$roll_open_mean_491520m)
#10m
ds$rom_10_3m <- (ds$roll_open_mean_10m/ds$roll_open_mean_3m)
ds$rom_10_5m <- (ds$roll_open_mean_10m/ds$roll_open_mean_5m)
ds$rom_10_15m <- (ds$roll_open_mean_10m/ds$roll_open_mean_15m)
I have a data frame with 40+ variables with 6 million rows, I have attached a smaller example data frame below.
Thanks in advance!
Charlie
X1 X2 X3 X4 X5 X6 X7 X8 X9 X10
1 57 77 48 8 31 43 47 13 26 88
2 25 75 86 77 4 65 5 49 31 57
3 91 90 42 69 82 33 56 99 47 39
4 35 96 86 77 67 77 20 17 77 92
5 6 100 50 62 16 31 0 39 72 4
6 90 34 74 89 71 37 73 45 24 28
7 24 22 92 13 57 97 32 2 12 80
8 74 59 49 2 97 100 15 37 15 67
9 43 38 66 97 8 20 85 25 97 67
10 82 4 56 40 42 46 44 98 98 76
11 60 68 92 99 81 92 78 59 23 81
12 22 57 37 100 7 1 89 41 40 56
13 69 13 1 82 89 45 83 24 71 29
14 8 14 66 48 94 8 20 3 28 63
15 26 70 56 62 9 34 11 86 71 64
16 7 55 15 100 91 89 46 74 98 14
17 29 68 19 66 83 29 84 76 90 45
18 27 76 6 48 17 28 8 7 52 37
19 68 58 51 75 60 57 74 46 98 93
20 15 15 89 55 23 3 3 8 32 37
21 78 49 57 48 96 89 4 95 67 58
22 12 36 42 59 27 92 48 0 92 28
23 51 17 77 61 84 53 46 22 27 36
24 40 84 83 35 19 13 80 78 96 87
25 44 80 25 72 43 17 74 70 52 36
26 14 61 63 82 16 47 32 93 19 84
27 93 19 28 62 74 1 85 65 50 9
28 80 62 6 58 48 97 97 18 65 43
29 12 58 95 79 37 89 89 83 22 85
30 57 73 22 88 99 63 58 87 90 66
As #27 ϕ 9 suggested in the comments you should use that lapply solution.
With this, you also create a unique dataframe with correct names
l <- lapply(df, `/`, df)
l <- unlist(l, recursive = FALSE)
data.frame(l)
This question already has answers here:
Split a vector into chunks
(22 answers)
Closed 3 years ago.
How to cut the values (1 to 100) in a regular interval (25) and place them into 4 groups as below:
sdr <- c(1:100)
Group1: 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25
Group2: 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50
Group3: 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75
Group4: 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100
Any suggestion, please.
You could use split
sdr <- 1:100
split(sdr, rep(1:4, each = 25))
#$`1`
# [1] 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25
#
#$`2`
# [1] 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50
#
#$`3`
# [1] 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75
#
#$`4`
# [1] 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94
#[20] 95 96 97 98 99 100
This returns a list with 4 vector elements.
Also note that the c() around 1:100 is not necessary.
Or we can define the number of groups
ngroup <- 4
split(sdr, rep(1:ngroup, each = length(sdr) %/% ngroup))
giving the same result.
You can make a dataframe for your groups and then transpose using t:
df <- t(data.frame(Group1 = c(1:25), Group2 = c(26:50), Group3 = c(51:75), Group4 = c(76:100)))
This is a rather simple question: why is this code in R not printing numbers from 1 to 100, but jumps with the value of i? Is there a way to prevent this?
t <-5
for (i in 1:t){
print(20*(i-1)+1:20*i)
}
to get the question closed
t <-5
for (i in 1:t){
print(20*(i-1)+1:20)
}
#> [1] 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21
#> [1] 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42
#> [1] 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63
#> [1] 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84
#> [1] 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100
How do I set the width of a reprex output?
Say I have a code like this:
(x <- 1:100)
I get this with reprex::reprex(venue = "so")
(x <- 1:100)
#> [1] 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17
#> [18] 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34
#> [35] 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51
#> [52] 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68
#> [69] 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85
#> [86] 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100
How can I increase the width of the output to output something like this
[1] 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50
[51] 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100
Possible Solutions
One option that I have found but I find rather "un-tidy" is this (include options(width = ...) at the top of the code. But I don't want it to show up in the output, I'd prefer setting the width in the reprex-call.
options(width = 205)
(x <- 1:100)
#> [1] 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50
#> [51] 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100
reprex() allows for knitr's opts-chunk, but I can't get it working with reprex::reprex(venue = "so", opts_chunk = list(out.width = 205)) (which might be related to #421 as pointed out here (Long lines of text output))
Any better solutions?
reprex has a syntax for setting these options but not including them in the output markdown (see here for examples). In this case:
reprex({
#+ setup, include = FALSE
options(width=205)
#+ actual-reprex-code
(x <- 1:100)
}, venue = 'so')
outputs your desired format:
(x <- 1:100)
#> [1] 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50
#> [51] 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100
Created on 2018-09-21 by the reprex package (v0.2.1)
This question already has answers here:
Get a seq() in R with alternating steps
(6 answers)
Closed 6 years ago.
I want to use R to create the sequence of numbers 1:8, 11:18, 21:28, etc. through 1000 (or the closest it can get, i.e. 998). Obviously typing that all out would be tedious, but since the sequence increases by one 7 times and then jumps by 3 I'm not sure what function I could use to achieve this.
I tried seq(1, 998, c(1,1,1,1,1,1,1,3)) but it does not give me the results I am looking for so I must be doing something wrong.
This is a perfect case of vectorisation( recycling too) in R. read about them
(1:100)[rep(c(TRUE,FALSE), c(8,2))]
# [1] 1 2 3 4 5 6 7 8 11 12 13 14 15 16 17 18 21 22 23 24 25 26 27 28 31 32
#[27] 33 34 35 36 37 38 41 42 43 44 45 46 47 48 51 52 53 54 55 56 57 58 61 62 63 64
#[53] 65 66 67 68 71 72 73 74 75 76 77 78 81 82 83 84 85 86 87 88 91 92 93 94 95 96
#[79] 97 98
rep(seq(0,990,by=10), each=8) + seq(1,8)
You want to exclude numbers that are 0 or 9 (mod 10). So you can try this too:
n <- 1000 # upper bound
x <- 1:n
x <- x[! (x %% 10) %in% c(0,9)] # filter out (0, 9) mod (10)
head(x,80)
# [1] 1 2 3 4 5 6 7 8 11 12 13 14 15 16 17 18 21 22 23 24 25 26 27
# 28 31 32 33 34 35 36 37 38 41 42 43 44 45 46 47 48 51 52 53 54 55 56 57
# 58 61 62 63 64 65 66 67 68 71 72 73 74 75 76 77 78 81 82 83 84 85
# 86 87 88 91 92 93 94 95 96 97 98
Or in a single line using Filter:
Filter(function(x) !((x %% 10) %in% c(0,9)), 1:100)
# [1] 1 2 3 4 5 6 7 8 11 12 13 14 15 16 17 18 21 22 23 24 25 26 27 28 31 32 33 34 35 36 37 38 41 42 43 44 45 46 47 48 51 52 53 54 55 56 57
# [48] 58 61 62 63 64 65 66 67 68 71 72 73 74 75 76 77 78 81 82 83 84 85 86 87 88 91 92 93 94 95 96 97 98
With a cycle: for(value in c(seq(1,991,10))){vector <- c(vector,seq(value,value+7))}