Hi I have a dataframe as such.
#Data
df2 <- structure(list(Date = structure(c(18502, 18503, 18504, 18505,
18506, 18507, 18508, 18509, 18510, 18511, 18512, 18513, 18514,
18515, 18516, 18517, 18518, 18519, 18520, 18521, 18522, 18523,
18524, 18525, 18526, 18527, 18528, 18529, 18530, 18531, 18532,
18533, 18534, 18535, 18536, 18537, 18538, 18539, 18540), class = "Date"),
Price = c(1490, 3604, 2003, -4004, 4247, -189008, 2506, 4044,
2604, 2204, -4316, 2190, 3137, 2694, 711, 4075, 1315, 454,
1660, 4306, 4032, 3201, 2980, 4474, 3044, 3267, 2573, 2784,
1497, 897, 4342, 4086, 3192, 3634, 380, 2293, 3478, 1190,
1619)), class = "data.frame", row.names = c(NA, -39L))
what I want is to find the absolute-value max and not just the max. For example, if the previous nth row is set to 4 then row 7 would be -189008. I have tried two functions but both do not have the option to return the absolute value.
df2 %>%
mutate(abshigh = slide_dbl(Price, max, .before = 4, .complete = TRUE))
zoo::rollmaxr (df2$Price,k=4,fill=NA)
above function outputs
Date Price abshigh
1 2020-08-28 1490 NA
2 2020-08-29 3604 NA
3 2020-08-30 2003 NA
4 2020-08-31 -4004 NA
5 2020-09-01 4247 4247
6 2020-09-02 -189008 4247
7 2020-09-03 2506 4247
8 2020-09-04 4044 4247
9 2020-09-05 2604 4247
10 2020-09-06 2204 4044
11 2020-09-07 -4316 4044
12 2020-09-08 2190 4044
13 2020-09-09 3137 3137
14 2020-09-10 2694 3137
15 2020-09-11 711 3137
16 2020-09-12 4075 4075
17 2020-09-13 1315 4075
18 2020-09-14 454 4075
19 2020-09-15 1660 4075
20 2020-09-16 4306 4306
21 2020-09-17 4032 4306
22 2020-09-18 3201 4306
23 2020-09-19 2980 4306
24 2020-09-20 4474 4474
25 2020-09-21 3044 4474
26 2020-09-22 3267 4474
27 2020-09-23 2573 4474
28 2020-09-24 2784 4474
29 2020-09-25 1497 3267
30 2020-09-26 897 3267
31 2020-09-27 4342 4342
32 2020-09-28 4086 4342
33 2020-09-29 3192 4342
34 2020-09-30 3634 4342
35 2020-10-01 380 4342
36 2020-10-02 2293 4086
37 2020-10-03 3478 3634
38 2020-10-04 1190 3634
39 2020-10-05 1619 3478
this is incorrect since row seven should return -189008 instead of 4247 thanks.
CodePudding user response：
One option is to get the slide max on the absolute 'Price' and then change the sign by matching the absolute values of 'Price' (potential bug alert!)
library(dplyr)
library(slider)
df2 %>%
mutate(abshigh = slide_dbl(abs(Price), max, .before = 4, .complete = TRUE),
abshigh = abshigh * sign(Price[match(abshigh, abs(Price))]))
-output
Date Price abshigh
1 2020-08-28 1490 NA
2 2020-08-29 3604 NA
3 2020-08-30 2003 NA
4 2020-08-31 -4004 NA
5 2020-09-01 4247 4247
6 2020-09-02 -189008 -189008
7 2020-09-03 2506 -189008
8 2020-09-04 4044 -189008
9 2020-09-05 2604 -189008
10 2020-09-06 2204 -189008
11 2020-09-07 -4316 -4316
12 2020-09-08 2190 -4316
13 2020-09-09 3137 -4316
14 2020-09-10 2694 -4316
15 2020-09-11 711 -4316
16 2020-09-12 4075 4075
17 2020-09-13 1315 4075
18 2020-09-14 454 4075
19 2020-09-15 1660 4075
20 2020-09-16 4306 4306
21 2020-09-17 4032 4306
22 2020-09-18 3201 4306
23 2020-09-19 2980 4306
24 2020-09-20 4474 4474
25 2020-09-21 3044 4474
26 2020-09-22 3267 4474
27 2020-09-23 2573 4474
28 2020-09-24 2784 4474
29 2020-09-25 1497 3267
30 2020-09-26 897 3267
31 2020-09-27 4342 4342
32 2020-09-28 4086 4342
33 2020-09-29 3192 4342
34 2020-09-30 3634 4342
35 2020-10-01 380 4342
36 2020-10-02 2293 4086
37 2020-10-03 3478 3634
38 2020-10-04 1190 3634
39 2020-10-05 1619 3478
Or another option is to use row_number() i.e. sequence as input to slide_dbl and extract the values of 'Price' based on the sequence, get the max element index of absolute values with which.max to subset the 'Price' subset (This would be more correct compared to the previous solution as match can have undesirable consequence on the absolute values if there are both negative and positive elements with same value
df2 %>%
mutate(abshigh = slide_dbl(row_number(),
.f = ~ Price[.x][which.max(abs(Price[.x]))], .before = 4, .complete = TRUE))
-output
Date Price abshigh
1 2020-08-28 1490 NA
2 2020-08-29 3604 NA
3 2020-08-30 2003 NA
4 2020-08-31 -4004 NA
5 2020-09-01 4247 4247
6 2020-09-02 -189008 -189008
7 2020-09-03 2506 -189008
8 2020-09-04 4044 -189008
9 2020-09-05 2604 -189008
10 2020-09-06 2204 -189008
11 2020-09-07 -4316 -4316
12 2020-09-08 2190 -4316
13 2020-09-09 3137 -4316
14 2020-09-10 2694 -4316
15 2020-09-11 711 -4316
16 2020-09-12 4075 4075
17 2020-09-13 1315 4075
18 2020-09-14 454 4075
19 2020-09-15 1660 4075
20 2020-09-16 4306 4306
21 2020-09-17 4032 4306
22 2020-09-18 3201 4306
23 2020-09-19 2980 4306
24 2020-09-20 4474 4474
25 2020-09-21 3044 4474
26 2020-09-22 3267 4474
27 2020-09-23 2573 4474
28 2020-09-24 2784 4474
29 2020-09-25 1497 3267
30 2020-09-26 897 3267
31 2020-09-27 4342 4342
32 2020-09-28 4086 4342
33 2020-09-29 3192 4342
34 2020-09-30 3634 4342
35 2020-10-01 380 4342
36 2020-10-02 2293 4086
37 2020-10-03 3478 3634
38 2020-10-04 1190 3634
39 2020-10-05 1619 3478