我正在研究作物生长的热需求。我有一个表，其中包含 6 个月时间段内的累积温度。示例如下：
date temp cum_temp
1: 2020-03-01 9.339748 9.339748
2: 2020-03-02 23.860849 33.200597
3: 2020-03-03 12.860331 46.060928
4: 2020-03-04 26.607505 72.668432
5: 2020-03-05 28.273551 100.941984
6: 2020-03-06 2.321138 103.263122
7: 2020-03-07 16.315059 119.578181
8: 2020-03-08 26.880152 146.458334
9: 2020-03-09 16.991615 163.449949
10: 2020-03-10 14.241827 177.691776
11: 2020-03-11 28.748167 206.439943
12: 2020-03-12 14.146691 220.586634
13: 2020-03-13 20.649548 241.236182
14: 2020-03-14 17.606369 258.842551
15: 2020-03-15 3.984816 262.827367
然后，我还有一张表格，其中列出了作物生长阶段及其热需求（即达到每个阶段所需的热阈值）：
growth_stage thermal_req
1: VE 120
2: V2 200
3: V3 350
4: V5-V6 475
5: V7-V9 610
6: R2 1660
7: R4 1925
8: R5 2450
9: R6 2700
根据这些表格，我需要两个结果：
growth_stage thermal_req date_reached
1: VE 120 2020-03-08
2: V2 200 2020-03-11
3: V3 350 2020-03-21
4: V5-V6 475 2020-03-26
5: V7-V9 610 2020-04-03
6: R2 1660 2020-06-14
7: R4 1925 2020-06-30
8: R5 2450 2020-08-06
9: R6 2700 2020-08-23
date temp cum_temp growth_stage
1: 2020-03-01 9.339748 9.339748 NA
2: 2020-03-02 23.860849 33.200597 NA
3: 2020-03-03 12.860331 46.060928 NA
4: 2020-03-04 26.607505 72.668432 NA
5: 2020-03-05 28.273551 100.941984 NA
6: 2020-03-06 2.321138 103.263122 NA
7: 2020-03-07 16.315059 119.578181 NA
8: 2020-03-08 26.880152 146.458334 VE
9: 2020-03-09 16.991615 163.449949 VE
10: 2020-03-10 14.241827 177.691776 VE
11: 2020-03-11 28.748167 206.439943 V2
12: 2020-03-12 14.146691 220.586634 V2
13: 2020-03-13 20.649548 241.236182 V2
14: 2020-03-14 17.606369 258.842551 V2
15: 2020-03-15 3.984816 262.827367 V2
16: 2020-03-16 27.094924 289.922291 V2
17: 2020-03-17 8.136544 298.058835 V2
18: 2020-03-18 2.219726 300.278562 V2
19: 2020-03-19 10.509701 310.788263 V2
20: 2020-03-20 28.680606 339.468868 V2
21: 2020-03-21 26.796640 366.265509 V3
22: 2020-03-22 21.091299 387.356807 V3
23: 2020-03-23 19.574698 406.931505 V3
24: 2020-03-24 29.833824 436.765328 V3
25: 2020-03-25 20.015468 456.780797 V3
26: 2020-03-26 21.547384 478.328180 V5-V6
27: 2020-03-27 16.777915 495.106095 V5-V6
28: 2020-03-28 18.230119 513.336214 V5-V6
29: 2020-03-29 9.385632 522.721846 V5-V6
30: 2020-03-30 5.266296 527.988142 V5-V6
31: 2020-03-31 28.927703 556.915844 V5-V6
32: 2020-04-01 27.166672 584.082517 V5-V6
33: 2020-04-02 21.030453 605.112970 V5-V6
34: 2020-04-03 24.068555 629.181525 V5-V6
35: 2020-04-04 1.713797 630.895322 V5-V6
36: 2020-04-05 14.856083 645.751405 V5-V6
37: 2020-04-06 22.995327 668.746732 V5-V6
38: 2020-04-07 7.275830 676.022562 V5-V6
39: 2020-04-08 10.227249 686.249811 V5-V6
40: 2020-04-09 7.717148 693.966959 V5-V6
date temp cum_temp growth_stage
实现这些成果的最佳方法是什么？
用于重现此问题的数据：
# load required packages
library(data.table)
# generate data
dates <- seq(as.Date("2020-03-01"), as.Date("2020-08-31"), by="days")
set.seed(123); temps <- runif(length(dates), min=1, max=30)
dat <- data.table(date=dates,
temp=temps)
# cumulative sum
dat$cum_temp <- cumsum(dat$temp)
# table with growth stage thermal requirements
sum_req <- data.table(growth_stage=c("VE","V2","V3","V5-V6","V7-V9","R2","R4","R5","R6"),
thermal_req=c(120,200,350,475,610,1660,1925,2450,2700))
可能是更好的方法，但如果日期有序且温度总是升高，则可能很简单：
dat$growth_stage <- NA
for (i in 1:nrow(sum_req)) {
indices <- which(dat$cum_temp >= sum_req$thermal_req[i])
dat$growth_stage[indices] <- sum_req$growth_stage[i]
}
查找数据集中哪些行的温度高于阈值， 然后将相应的生长阶段写入该栏。一开始就很多；当我们进入不同的成长阶段时，此列会被覆盖。
再次，在这里工作，可能不能很好地概括