将两列中的数据相乘并将其添加到现有的第三列

Question

I have a data Frame with two columns containing plant leaf lenght and width data.

For the year 2019 I have a mix of data. Some Data points only have length and area measurements. Some other data points have all three measurements.

With this data I was able to calculate a conversion factor. In 2020 I only have the length and width measurements. With the conversion factor I want to calculate the area for the year 2020 and add it to the leaf area column without overwriting any of the area measurements in 2019

df_all <- df_all%>% mutate(rep_leaf_length*rep_leaf_width * 0.790590)

This was my first starting point before I realized I do not know how to get where I want.

Do you guys have an Idea how to do the multiplication and add the result to the existing column but only for 2020 or if NA's are in the area column and not already existing area measurements.

Year  rep_leaf_length   rep_leaf_width  rep_leaf_area
2019           37.400               NA             NA
2019           21.036            8.080        132.914
2019           29.147            2.331             NA
2020           16.600             4.00             NA
2020           21.600              2.2             NA

Thanks a lot Jan

Answer 1

I think you mean that you want to infer the (unmeasured) leaf area from 2020, using the (measured) leaf length and leaf width from that year. However, the leaf area isn't a simple product of width and length, since leaves aren't rectangular. Fortunately, you have some observations from 2019 where length, width and area were all measured. That means if you compare the length * width to the actual area for the complete 2019 observations, you will get a ratio of actual area to (length * width). Since the leaves are presumably of relatively fixed shape, this ratio can be used to multiply the (length * width) values from 2020 to give an estimated area.

Assuming I have interpreted your intentions correctly, we can work out the ratio of actual area to (width * length) in 2019 like this:

library(dplyr)

ratio <- df_all %>% 
  filter(Year == 2019) %>%
  filter(complete.cases(.)) %>%
  summarize(ratio = mean(rep_leaf_area / (rep_leaf_length * rep_leaf_width))) %>%
  unlist()

ratio
#>    ratio 
#> 0.781981

And we can use the ratio like this:

df_all %>% 
  mutate(rep_leaf_area = ifelse(Year == 2020,
                                rep_leaf_length * rep_leaf_width * ratio,
                                rep_leaf_area))
#>   Year rep_leaf_length rep_leaf_width rep_leaf_area
#> 1 2019          37.400             NA            NA
#> 2 2019          21.036          8.080     132.91400
#> 3 2019          29.147          2.331            NA
#> 4 2020          16.600          4.000      51.92354
#> 5 2020          21.600          2.200      37.15974

Note that this does not affect 2019's area measurements.

---

Data

df_all <- structure(list(Year = c(2019L, 2019L, 2019L, 2020L, 2020L), 
           rep_leaf_length = c(37.4, 21.036, 29.147, 16.6, 21.6), 
           rep_leaf_width = c(NA, 8.08, 2.331, 4, 2.2), 
           rep_leaf_area = c(NA, 132.914, NA, NA, NA)), 
           class = "data.frame", row.names = c(NA, -5L))

Answer 2

Use an index. Calculate areas of indexed rows and assign them.

index <- dat$Year %in% 2020
areas <- apply(dat[index, 2:3], 1, prod)
dat[index, 4] <- areas  
#   Year rep_leaf_length rep_leaf_width rep_leaf_area
# 1 2019          37.400             NA            NA
# 2 2019          21.036          8.080       132.914
# 3 2019          29.147          2.331            NA
# 4 2020          16.600          4.000        66.400
# 5 2020          21.600          2.200        47.520

---

Data:

dat <- structure(list(Year = c(2019L, 2019L, 2019L, 2020L, 2020L), rep_leaf_length = c(37.4, 
21.036, 29.147, 16.6, 21.6), rep_leaf_width = c(NA, 8.08, 2.331, 
4, 2.2), rep_leaf_area = c(NA, 132.914, NA, NA, NA)), class = "data.frame", row.names = c(NA, 
-5L))