R: Extracting values from a dataframe when sequential row values differ - r

Question

I have a data.frame df with 2 columns:
A contains positive values.
B contains values (either zero or a positive values).
I wish to generate a new data.frame (or vector) (of unknown length) containing
the values from df[i+1, A], ONLY when df[i, B] == 0 & df[i + 1, B] != 0.
I can visualize how to do this by sequentially stepping through the data.frame using a loop, but this will take forever with >200,000 rows. What is the vectorized solution to a problem like this that requires arithmetic on sequential rows of a vector or data.frame?
Data is in this form:
A B
1 5 5
2 10 3
3 15 0
4 20 6
5 25 5
6 30 0
7 35 0
8 40 11
9 45 3
etc etc etc
I'd then like to extract the values of A from row 4 (A = 20) and row 8 (A = 40) etc.

Answer 1

You could use
df$A[-1][diff(df$B != 0) > 0]
[1] 20 40
The idea is as follows. First, given a vector c(1, 2), one way to extract 2 is of course c(1, 2)[2]. Another way is c(1, 2)[c(FALSE, TRUE)], i.e. you might subset a vector by using a logical vector.
After you edited your question, I see that we are no longer interested in the first row of df, so that is why I start with df$A[-1]. Then one way that is longer and very likely less efficient, but follows more readable logic, is
df$A[-1][df$B[-nrow(df)] == 0 & df$B[-1] != 0]
where df$B[-1] != 0 returns a logical vector corresponding to your condition df [ i+1, B ] != 0. Then df$B[-nrow(df)] == 0 returns another logical vector corresponding to df [ i, B ]==0. Then operator & performs element-wise AND operation, returns the final logical vector and gives the result.
Now diff(df$B != 0) > 0 is just a trickier way to write the same thing. df$B != 0 gives a logical vector. Then while performing diff(df$B != 0) we are taking differences of 1's (correspond to entries TRUE) and 0's (correspond to FALSE). For example, c(0, 1) != 0 gives c(FALSE, TRUE), which can be seen as c(0, 1), and then diff gives 1. So, we have ones in diff(df$B != 0) where entry 0 is followed by some nonzero (in your case - positive) number. To use these results for subsetting df$A[-1] we obtain the final logical vector with diff(df$B != 0) > 0.

Answer 2

Another option comes through 'dplyr' with the following code:
library(dplyr)
df %>% filter(B != 0 & lag(B, 1) == 0)
This uses the dataframe and leaves rows where B doesn't equal 0 and the prior B does equal zero. This does return columns A and B. If you want to see only certain columns add %>% select(...) with the argument being variables separated by commas.

Answer 3

My example (adds together sequential values from two vectors):
> i1=c(1:100)
> i2=c(100:1)
> i3=i1[-length(i1)]+i2[-1]
> i3
[1] 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100
[55] 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100 100
My example (adds together sequential values from a single vector):
> i1=c(1:100)
> i2=i1[-length(i1)]+i1[-1]
> i2
[1] 3 5 7 9 11 13 15 17 19 21 23 25 27 29 31 33 35 37 39 41 43 45 47 49 51 53 55 57 59 61 63 65 67 69 71 73 75 77 79 81 83 85 87 89 91 93 95 97 99 101 103 105 107 109
[55] 111 113 115 117 119 121 123 125 127 129 131 133 135 137 139 141 143 145 147 149 151 153 155 157 159 161 163 165 167 169 171 173 175 177 179 181 183 185 187 189 191 193 195 197 199