如何检测非 NA 值的最早实例？

Question

I have a data frame which lists events against the dates they occurred.

How can I record the first date a non- NA value appears in the dataset?

For example, given this dataset:

Colour       Date  
<chr>        <date>    
Blue         2021-06-29
NA           2021-03-10
NA           2021-04-02
Amber        2021-04-09
Blue         2021-06-21
NA           2021-03-09
Blue         2021-04-11
Amber        2021-05-21
NA           2021-02-17

How can I find the first/earliest time a non- NA value appears in terms of date/history?

In this example, the result would be Amber, 2021-04-09 as the values that occur before it in time are all NA up until this point.

Answer 1

If we need the first non-NA by 'Date', arrange first and then slice the first row

library(dplyr)
df1 %>%
   arrange(is.na(Colour), Date) %>% 
   slice_head(n = 1)  

-output

Colour       Date
1  Amber 2021-04-09

data

df1 <- structure(list(Colour = c("Blue", NA, "Amber", "Blue", NA, "Blue", 
"Amber", NA), Date = structure(c(18807, 18696, 18726, 18799, 
18695, 18728, 18768, 18675), class = "Date")), row.names = c(NA, 
-8L), class = "data.frame")

Answer 2

I guess you can try the code below

> subset(df[order(df$Date), ], min(which(!is.na(Colour))) == seq_along(Colour))
  Colour       Date
3  Amber 2021-04-09

Answer 3

Order the date, drop NA values and select the first row.

library(dplyr)

df1 %>% arrange(Date) %>% na.omit() %>% slice(1L)

#  Colour       Date
#1  Amber 2021-04-09

In base R -

na.omit(df1[order(df1$Date), ])[1, ]

Answer 4

Without using any library:

df1 = df[order(df$Date),] # orders by date
df1[which(!is.na(df1$Colour))[1],] # pick the first non-empty colour