Related
Participants in an experiment took a test that has a rule that says "once a participant has gotten 6 items wrong in a window of 8 items, you stop running the test". However, some experimenters kept testing past this point. I now need to find a way in which I can automatically see where the test should have been stopped, and change all values following the end to 0 (= item wrong). I am not even sure if this is something that can be done in R.
To be clear, I would like to go row by row (which are the participants) and once there are six 0s in a given window of 8 columns (items), I would need all values after the sixth 0 to be 0 too.
While the reproducible data is below, here is a visualization of what I would need, where the blue cells are the ones that should change to 0:
Pre-changes
Post-changes
Reproducible data:
structure(list(Participant_ID = c("E01P01", "E01P02", "E01P03",
"E01P04", "E01P05", "E01P06", "E01P07", "E01P08", "E02P01", "E02P02"
), A2 = c(1, 1, 1, 0, 0, 1, 1, 1, 1, 1), A3 = c(1, 1, 0, 0, 0,
1, 0, 0, 0, 0), B1 = c(1, 1, 1, 0, 0, 1, 0, 0, 1, 1), B2 = c(1,
1, 1, 1, 1, 1, 0, 0, 0, 1), C3 = c(1, 0, 0, 1, 0, 1, 0, 0, 0,
1), C4 = c(1, 0, 0, 0, 0, 1, 0, 0, 1, 1), D1 = c(1, 0, 0, 0,
0, 1, 0, 0, 0, 0), D3 = c(1, 1, 1, 1, 0, 0, 1, 0, 0, 1), E1 = c(1,
0, 0, 0, 0, 1, 0, 0, 0, 1), E3 = c(1, 1, 0, 1, 0, 1, 0, 0, 0,
0), F1 = c(1, 0, 0, 0, 1, 0, 0, 1, 0, 0), F4 = c(1, 1, 1, 1,
0, 1, 0, 1, 1, 0), G1 = c(1, 0, 0, 0, 0, 1, 0, 0, 0, 1), G2 = c(0,
0, 0, 0, 1, 1, 1, 0, 1, 1)), row.names = c(NA, -10L), class = c("tbl_df",
"tbl", "data.frame"))
Any help is highly appreciated!
Here is a solution that involves some pivoting, rollsum, cumsum, if_else logic, then pivoting back. Let me know if it works.
library(tidyverse)
library(zoo)
structure(list(Participant_ID = c("E01P01", "E01P02", "E01P03",
"E01P04", "E01P05", "E01P06", "E01P07", "E01P08", "E02P01", "E02P02"
), A2 = c(1, 1, 1, 0, 0, 1, 1, 1, 1, 1), A3 = c(1, 1, 0, 0, 0,
1, 0, 0, 0, 0), B1 = c(1, 1, 1, 0, 0, 1, 0, 0, 1, 1), B2 = c(1,
1, 1, 1, 1, 1, 0, 0, 0, 1), C3 = c(1, 0, 0, 1, 0, 1, 0, 0, 0,
1), C4 = c(1, 0, 0, 0, 0, 1, 0, 0, 1, 1), D1 = c(1, 0, 0, 0,
0, 1, 0, 0, 0, 0), D3 = c(1, 1, 1, 1, 0, 0, 1, 0, 0, 1), E1 = c(1,
0, 0, 0, 0, 1, 0, 0, 0, 1), E3 = c(1, 1, 0, 1, 0, 1, 0, 0, 0,
0), F1 = c(1, 0, 0, 0, 1, 0, 0, 1, 0, 0), F4 = c(1, 1, 1, 1,
0, 1, 0, 1, 1, 0), G1 = c(1, 0, 0, 0, 0, 1, 0, 0, 0, 1), G2 = c(0,
0, 0, 0, 1, 1, 1, 0, 1, 1)), row.names = c(NA, -10L), class = c("tbl_df",
"tbl", "data.frame")) %>%
as_tibble() %>%
pivot_longer(-1) %>%
group_by(Participant_ID) %>%
mutate(running_total = zoo::rollsumr(value==0, k = 8, fill = 0),
should_terminate = cumsum(running_total >= 6),
value = if_else(should_terminate > 0, 0, value)) %>%
ungroup() %>%
select(Participant_ID, name, value) %>%
pivot_wider(names_from = name, values_from = value)
Consider this data:
df <- structure(list(V1 = c(1, 1, 1, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0,
0, 0, 0, 0, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 1, 1, 1,
1, 1, 1, 1, 1, 1, 1, 1, 0, 0, 0, 0), V2 = c(0, 0, 0, 0, 0, 0,
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 1, 1, 1, 1, 0, 0, 0, 0, 0,
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 1, 1, 1, 1, 1, 1, 0, 0), V3 = c(1,
1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 0, 0, 0, 0, 1, 1, 0, 1, 1,
1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,
1, 1, 1), V4 = c(1, 1, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
0, 0, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
0, 0, 0, 0, 1, 1, 1, 0, 0, 0), V5 = c(1, 1, 0, 0, 0, 0, 0, 0,
0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 1, 1), V6 = c(0,
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 1, 0, 0, 0,
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 1, 1, 1, 1,
1, 0, 0), V7 = c(0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
0, 0, 1, 1, 0, 0, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 1, 1,
1, 1, 1, 1, 0, 0, 0, 0, 0, 0), V8 = c(1, 1, 1, 0, 0, 0, 0, 1,
1, 0, 1, 1, 0, 0, 0, 0, 0, 0, 0, 1, 1, 1, 1, 1, 0, 0, 0, 0, 0,
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 1, 1, 1, 1, 0, 0), V9 = c(1,
1, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 1, 0, 0, 0,
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 1, 1, 1, 1,
1, 1, 1)), class = "data.frame", row.names = c(NA, -46L))
I would like to apply the following functions for all possible pairs in df.
Applying to the pair V1 and V2:
df$V1V2 <- (df$V1 * df$V2) + (1 - df$V1) * (1 - df$V2)
sum <- (sum(df$V1V2)/46)
df$VD <- (df$V1/sd(df$V1))
df$VI <- (df$V2/sd(df$V2))
est <- lm(df$VD ~ df$VI)
summary(est)
ndf <- data.frame(NA)
ndf$V1V2 <- summary(est)[["coefficients"]][, "Pr(>|t|)"][2]
Applying to the pair V1 and V3:
df$V1V3 <- (df$V1 * df$V3) + (1 - df$V1) * (1 - df$V3)
sum <- (sum(df$V1V3)/46)
df$VD <- (df$V1/sd(df$V1))
df$VI <- (df$V3/sd(df$V3))
est <- lm(df$VD ~ df$VI)
summary(est)
ndf$V1V3 <- summary(est)[["coefficients"]][, "Pr(>|t|)"][2]
I could apply this to all other pairs of V1 (i.e. V3, V4, V5, V6, V7, V8, and V9). Nonetheless, I am sure this is not the best approach.
UPDATE
This is how far I got:
dfV1 <- df[, c("V1", "V2", "V1", "V3", "V1", "V4", "V1", "V5", "V1", "V6", "V1", "V7", "V1", "V8", "V1", "V9")]
colnames(dfV1) <- c("V1", "V2", "V1", "V3", "V1", "V4", "V1", "V5", "V1", "V6", "V1", "V7", "V1", "V8", "V1", "V9")
sep <- lapply(seq(1, ncol(dfV1), by=2), function(i)
dfV1[i: pmin((i+1), ncol(dfV1))])
V1V2 <- sep[[1]]
V1V3 <- sep[[2]]
V1V4 <- sep[[3]]
V1V5 <- sep[[4]]
V1V6 <- sep[[5]]
V1V7 <- sep[[6]]
V1V8 <- sep[[7]]
V1V9 <- sep[[8]]
list_V1 <- tibble::lst(V1V2, V1V3, V1V4, V1V5, V1V6, V1V7, V1V8, V1V9)
library(dplyr)
my_func <- function(x) {
x <- x %>%
mutate(First = (x[,1] * x[,2] + (1 - x[,1] * (1 - x[,2]))),
VD = x[,1] / sd(x[,1]),
VI = x[,2] / sd(x[,2]))
}
res <- lapply(list_V1, my_func)
list2env(res, .GlobalEnv)
df.IC.V1 <- cbind.data.frame(V1V2$First, V1V3$First, V1V4$First, V1V5$First, V1V6$First, V1V7$First, V1V8$First, V1V9$First)
IC.all.V1 <- data.frame(colSums(df.IC.V1)/46)
I do not know how to apply this part to the list dfV1:
est <- lm(df$VD ~ df$VI)
summary(est)
ndf$V1V3 <- summary(est)[["coefficients"]][, "Pr(>|t|)"][2]
Avoid wide data and keep data long or tidy which helps in virtually every aspect of data science: aggregation, modeling, and plotting. Therefore, consider reshape (or tidy semantics of pivot_longer) to reformat data to long and with a generalized method, call by to run your model by each of the 9 different variables.
Finally, you may want to separate out results of this process in a new data frame to avoid repetition of values in original due to different lengths and especially since your model outputs two rows for intercept and variable with multiple columns.
long_df <- reshape(df, idvar="V1", varying=names(df)[-1],
times=names(df)[-1],
v.names="value", timevar="variable",
new.row.names=1:1E4, direction="long")
pairwise_model <- function(sub) {
sub$vpair <- (sub$V1 * sub$value) + (1 - sub$value) * (1 - sub$value)
sum_v <- (sum(sub$vpair)/46)
sub$VD <- (sub$V1/sd(sub$V1))
sub$VI <- (sub$value/sd(sub$value))
est <- lm(VD ~ VI, data=sub)
print(summary(est))
ndf <- data.frame(variable = sub$variable[[1]],
summary(est)[["coefficients"]])
return(ndf)
}
df_list <- by(long_df, long_df$variable, pairwise_model)
results_df <- do.call(rbind, df_list)
results_df
# variable Estimate Std..Error t.value Pr...t..
# V2.(Intercept) V2 0.70897157 0.1697361 4.1769050 1.376794e-04
# V2.VI V2 0.20546559 0.1475392 1.3926169 1.707334e-01
# V3.(Intercept) V3 0.00000000 0.4324256 0.0000000 1.000000e+00
# V3.VI V3 0.29294628 0.1441419 2.0323471 4.818178e-02
# V4.(Intercept) V4 0.59719677 0.1461565 4.0860091 1.829775e-04
# V4.VI V4 0.47663445 0.1325296 3.5964390 8.118808e-04
# V5.(Intercept) V5 0.77259722 0.1604268 4.8158874 1.766754e-05
# V5.VI V5 0.13627939 0.1493492 0.9124883 3.664845e-01
# V6.(Intercept) V6 0.68720490 0.1557572 4.4120255 6.533736e-05
# V6.VI V6 0.31399751 0.1431310 2.1937768 3.357833e-02
# V7.(Intercept) V7 0.57392936 0.1516681 3.7841150 4.627413e-04
# V7.VI V7 0.46128020 0.1337587 3.4486009 1.253140e-03
# V8.(Intercept) V8 0.90047538 0.1869717 4.8161045 1.765498e-05
# V8.VI V8 -0.09345783 0.1500959 -0.6226544 5.367256e-01
# V9.(Intercept) V9 0.60871296 0.1645124 3.7001029 5.960399e-04
# V9.VI V9 0.35598290 0.1408800 2.5268512 1.517916e-02
Online Demo
I have two logical vectors which look like this:
x = c(0, 0, 0, 1, 1, 1, 1, 0, 0, 0, 0, 0, 0)
y = c(0, 0, 0, 0, 0, 1, 1, 1, 1, 1, 1, 0, 0)
I would like to count the intersections between ranges of consecutive values. Meaning that consecutive values (of 1s) are handled as one range. So in the above example, each vector contains one range of 1s and these ranges intersect only once.
Is there any R package for range intersections which could help here?
I think this should work (calling your logical vectors x and y):
sum(rle(x & y)$values)
A few examples:
x = c(0, 0, 0, 1, 1, 1, 1, 0, 0, 0, 0, 0, 0)
y = c(0, 0, 0, 0, 0, 1, 1, 1, 1, 1, 1, 0, 0)
sum(rle(x & y)$values)
# [1] 1
x = c(1, 1, 0, 1, 1, 1, 1, 0, 0, 0, 0, 0, 0)
y = c(0, 1, 1, 0, 0, 1, 1, 1, 1, 1, 1, 0, 0)
sum(rle(x & y)$values)
# [1] 2
x = c(1, 1, 0, 1, 1, 1, 1, 0, 0, 0, 1, 1, 0)
y = c(0, 1, 1, 0, 0, 1, 1, 1, 1, 1, 1, 0, 0)
sum(rle(x & y)$values)
# [1] 3
By way of explanation, x & y gives the intersections on a per-element level, rle collapses runs of adjacent intersections, and sum counts.
This issue is quite tricky to explain but I am sure some of you already faced it.
So I have two matrix.
Matrix 1 (mat 1) and
Matrix 2 (mat 2)
What I want to do is to record in a third matrix (mat3) the value of mat2, after checking for matrix 1, but with a LAG. Let me explain.
After the value 1 in matrix 1, I want to check if matrix 2 as a 1 too but within the range of a certain LAG, for example, 1 or 2 episodes after (column).
For example, row number 4 has a 1 in matrix 1 at the 6th column.
So I want to check if in matrix 2 for row number 4 it has a 1 directly after or after 2 or 3 more columns.
Do you understand the idea ?
mat1 = structure(c(0, 0, 0, 0, 1, 0, 0, 0, 0, 1, 0, 0, 0, 0, 1, 0, 0,
0, 0, 1, 0, 0, 0, 0, 1, 1, 0, 0, 0, 1, 0, 0, 0, 0, 1, 1, 1, 0,
0, 1, 0, 0, 0, 0, 1, 1, 1, 1, 0, 1, 0, 0, 1, 1, 1, 1, 1, 1, 0,
1, 0, 0, 1, 1, 1, 1, 1, 1, 0, 1, 0, 0, 1, 1, 1, 1, 1, 1, 0, 1,
0, 0, 1, 1, 1, 1, 1, 1, 0, 1, 0, 0, 1, 1, 1, 1, 1, 1, 0, 1, 0,
0, 1, 1, 1, 1, 1, 1, 0, 1, 0, 0, 1, 1, 1, 1, 1, 1, 1, 1, 0, 0,
1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,
1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,
1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,
1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,
1, 1, 1, 1), .Dim = c(10L, 21L), .Dimnames = list(NULL, c("wit5.020",
"wit5.021", "wit5.022", "wit5.023", "wit5.024", "wit5.025", "wit5.026",
"wit5.027", "wit5.028", "wit5.029", "wit5.030", "wit5.031", "wit5.032",
"wit5.033", "wit5.034", "wit5.035", "wit5.036", "wit5.037", "wit5.038",
"wit5.039", "wit5.040")))
mat2 = structure(c(0, 0, 0, 0, 1, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1, 0, 0,
0, 0, 0, 0, 0, 0, 0, 1, 1, 0, 0, 0, 0, 0, 0, 0, 0, 1, 1, 1, 0,
0, 0, 0, 0, 0, 0, 1, 1, 1, 1, 0, 0, 0, 0, 1, 1, 1, 1, 1, 1, 0,
0, 0, 0, 1, 1, 1, 1, 1, 1, 0, 0, 0, 0, 1, 1, 1, 1, 1, 1, 0, 0,
0, 0, 1, 1, 1, 1, 1, 1, 0, 0, 0, 0, 1, 1, 1, 1, 1, 1, 0, 0, 0,
0, 1, 0, 1, 1, 1, 1, 0, 0, 0, 0, 1, 1, 1, 1, 1, 1, 0, 1, 0, 0,
1, 1, 1, 1, 1, 1, 0, 1, 0, 0, 1, 1, 1, 1, 1, 1, 1, 1, 0, 0, 1,
0, 1, 1, 1, 1, 1, 1, 0, 0, 1, 0, 1, 0, 1, 1, 1, 1, 0, 0, 1, 0,
1, 0, 1, 1, 0, 1, 1, 0, 0, 0, 1, 0, 1, 1, 0, 1, 1, 0, 0, 0, 1,
0, 1, 1, 0, 1, 1, 0, 0, 0, 1, 0, 1, 1, 0, 1, 1, 0, 0, 0, 0, 0,
0, 1, 0, 1), .Dim = c(10L, 21L))
So mat3 - where I want to store the result of the check
mat3 = matrix(0, nrow = nrow(mat1), ncol = ncol(mat1))
So here is an example of a possible loop
in order to check the LAG - this loop doesn't work but it could give you an idea maybe of the solution.
I am not sure where to introduce the lag. I thought maybe in the i, but I am not sure.
for(j in 1:ncol(mat1)){
for(i in 1:nrow(mat1)){
if( mat1[i,j] == 1 & mat2[i,j] == 1 | mat2[i+1,j] == 1 | mat2[i+2,j] == 1) # lag here
{mat[i,j] <- 1}
else
{mat[i,j] <- 0}
}
}
Any ideas are very welcome.
Here's a simple way to do it:
lag <- 3 # or whatever lag you want
nr <- nrow(mat1)
nc <- ncol(mat1)
mat3 <- matrix(0, ncol=nc, nrow=nr)
for (r in 1:nr) {
for (c in 1:nc) {
if (mat1[r,c] == 1 && any(mat2[r,c:min(c+lag,nc)] == 1))
mat3[r,c] <- 1
}
}
Note the use of mat2[r,c:min(c+lag,nc)]. This selects all elements from current column c up through column c + lag, but it makes sure not to go past nc (the total number of columns). That is, this code is used to avoid an out-of-bounds error.
There's probably a faster, more vectory way of doing this, but the above code should work.
I everybody I remove my last post to make a reproducible exmaple of my problem. I am working with the next to data frames a1 (dput structure):
structure(list(r04_numero_operacion = c("0050475725", "0050490602",
"0050491033", "0050496386", "0050518985", "0050630090", "0050631615",
"0060235906", "0060238732", "0060241333", "0060244391", "0060245813",
"0060260056", "0060266356", "0800041441", "0800054041", "0800055382",
"0800058554", "2020200062", "2020200073", "CAR1010001706000",
"CAR1010001795000", "CAR1010001803000", "CAR1010001871000", "CAR1010001962000",
"CAR1010002002000", "CAR1010002120000", "CAR1010002189000", "CAR1010002215000",
"CAR1010002250000"), perdida3 = c(523.12, 265.43, 8371.66, 5242.13,
4960.51, 8473.27, 3743.45, 1283.32, 2229.25, 8001.27, 8653.94,
3670.13, 4536.02, 8216.55, 2481.36, 288.94, 1637.28, 4566.89,
1573.63, 11217.92, 0, 0, 0, 0, 0, 0, 0, 0, 9633.9, 0), Saldo = c(1,
1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 288.94, 1637.28, 4566.89,
1, 1, 481.59, 299.52, 258.13, 603.84, 231.61, 631.68, 220.6,
210.54, 1, 1224.44), Bvencida = c(0, 0, 0, 0, 0, 0, 0, 0, 0,
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 603.84, 0, 631.68,
0, 0, 0, 0), Cvencida = c(0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 1224.44),
Dvencida = c(0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0), vencida = c(1,
1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 288.94, 1637.28,
4566.89, 1, 1, 0, 0, 0, 603.84, 0, 631.68, 0, 0, 1, 1224.44
), V1 = c(1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,
1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1)), .Names = c("r04_numero_operacion",
"perdida3", "Saldo", "Bvencida", "Cvencida", "Dvencida", "vencida",
"V1"), codepage = 1252L, row.names = c(NA, 30L), class = "data.frame")
And a2 data frame (dput structure):
structure(list(r04_numero_operacion = c("0050475725", "0050490602",
"0050491033", "0050496386", "0050518985", "0050630090", "0050631615",
"0060235906", "0060238732", "0060241333", "0060244391", "0060245813",
"0060260056", "0060266356", "0800041441", "0800054041", "0800055382",
"0800058554", "2020200073", "CAR1010002002000", "CAR1010002189000",
"CAR1010002215000", "CAR1010002250000", "CAR1010002264000", "CAR1010002297000",
"CAR1010002401000", "CAR1010002412000", "CAR1010002436000", "CAR1010002529000",
"CAR1010002709000"), perdida3 = c(523.12, 265.43, 8371.66, 5242.13,
4960.51, 8473.27, 3743.45, 1283.32, 2229.25, 8001.27, 8653.94,
3670.13, 4536.02, 8216.55, 2481.36, 288.94, 1637.28, 4566.89,
11217.92, 0, 0, 9633.9, 0, 0, 0, 0, 0, 0, 0, 0), Saldo = c(1,
1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 288.94, 1637.28, 4566.89,
1, 317.72, 210.54, 1, 868.93, 242.91, 298.78, 120.63, 255.01,
357.68, 284.08, 308.83), Bvencida = c(0, 0, 0, 0, 0, 0, 0, 0,
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 317.72, 0, 0, 0, 0, 0, 0, 0,
0, 0, 0), Cvencida = c(0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
0, 0, 0, 0, 0, 0, 0, 0, 0, 868.93, 0, 0, 0, 0, 0, 0, 0), Dvencida = c(0,
0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0,
0, 0, 0, 0, 0, 0, 0, 0), vencida = c(1, 1, 1, 1, 1, 1, 1, 1,
1, 1, 1, 1, 1, 1, 1, 288.94, 1637.28, 4566.89, 1, 317.72, 0,
1, 868.93, 0, 0, 0, 0, 0, 0, 0), V2 = c(2, 2, 2, 2, 2, 2, 2,
2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2, 2,
2, 2)), .Names = c("r04_numero_operacion", "perdida3", "Saldo",
"Bvencida", "Cvencida", "Dvencida", "vencida", "V2"), class = "data.frame", row.names = c(NA,
30L))
My problem is when I use merge() and match() functions. merge() is more functional than match() related to add new variables by common one but when I use merge() I don't get the same result as match(). First I used merge() with a2 and a1 to create DF with the next code:
DF=merge(a2,a1,all.x=TRUE)
It added V1 variable from a1 to DF and I got this summary for DF$V1:
Min. 1st Qu. Median Mean 3rd Qu. Max. NA's
1 1 1 1 1 1 9
After I create a copy of a2 named DF and I made a match with r04_numero_operacion using this code to add V1 variable from a1 to a2:
a2$V1<-a1[match(a2$r04_numero_operacion,a1$r04_numero_operacion),"V1"]
It added `V1 to DF but the result is different to the merge() way. I got this summary for DF$V1 in match() solution:
Min. 1st Qu. Median Mean 3rd Qu. Max. NA's
1 1 1 1 1 1 7
My problem is I want to make the same I made with match() but using merge() function due to this function is more poweful than match(). Thanks for your help.
In using match(a2$r04_numero_operacion,a1$r04_numero_operacion) the a2$r04_numero_operacion values gets matched the coresponding column in a1 while in using merge(a2,a1,all.x=TRUE) the a1 all the matching columns get matched to the matching column names in a2. If you only match on the first column, the NA counts match up:
summary( merge(a2,a1,by=1,all.x=TRUE)$V1 )
Min. 1st Qu. Median Mean 3rd Qu. Max. NA's
1 1 1 1 1 1 7