I have a data frame (raw.Delta):
Time Open High Low Close
1 12/18/2022 6:02:29 PM 1 1 1 1
2 12/18/2022 6:04:53 PM 1 43 -17 7
3 12/18/2022 6:06:46 PM 7 130 -6 92
4 12/18/2022 6:09:01 PM 92 145 68 124
5 12/18/2022 6:11:30 PM 124 205 118 135
6 12/18/2022 6:14:41 PM 135 179 117 130
...that I am trying to call ATR on like so...
delta <- raw.Delta %>%
mutate(dATR=ATR(c("High", "Low",
"Close"), n=50, maType="SMA"))
and every time I get....
Error in `mutate()`:
! Problem while computing `dATR =
``ATR(c("High", "Low", "Close"), n = 50, maType = "SMA")`.
Caused by error in `HLC[-NROW(HLC), 3]`:
! subscript out of bounds``
...even if I try to do:
mutate(dATR=ifelse(nBar>50,ATR(c("High",
"Low", "Close"), n=50,
maType="SMA"),0))
it will throw the same error. I have also tried not specifying 'maType', I have tried simply using 'Close' instead of c("H","L","C"). I feel like I have tried every combination to get it to work. FYI this...
SMA(raw.Delta, n=50)
...works perfectly fine even though it also has a lookback period AND I don't even need to do a check on bar number to ensure there are enough bars in the series, it simply creates a new column in my df (when called by mutate) where the first 50 values are NA, which I believe is the expected behavior of ATR as well.
Anyway, I hope I have provided enough information for someone to be able to spot the issue. Thanks.
%>% passes the LHS as the first argument to ATR(). So your code is the same as ATR(raw.Delta, n = 50, maType = "SMA", c("High", "Low", "Close")), which isn't what you want.
This should work:
raw.Delta %>%
mutate(dATR = ATR(.[, c("High", "Low", "Close")], n = 50, maType = "SMA"))
# or use quantmod::HLC() to grab the high, low, close columns
raw.Delta %>% mutate(dATR = ATR(HLC(.), n = 50, maType = "SMA"))