Wie bestellen die Zeilen Informationen eines Datensatzes mit zwei Kriterien

Question

ich einen Datensatz haben, die Informationen über akademische Abschlüsse pro Jahr, wie folgt aus:

Year1 Deg_Year1 Year2 Deg_Year2 Year3 Deg_Year3 Year4 Deg_Year4 Year5 Deg_Year5 
2001 College 2004 Master  NA  NA  NA  NA  NA  NA 
2004 College 2004 Master  2010 PHD  NA  NA  NA  NA 
2006 Master  2006 College  NA  NA  NA  NA  NA  NA 
2016 Master  NA  NA   NA  NA  NA  NA  NA  NA 
2002 Master  2003 Master  2004 College 2004  Master NA  NA 
2014 Master  2017 PHD   NA  NA  NA  NA  NA  NA 

ich einen Datenrahmen erhalten möchten, die das Jahr enthält und der höchste akademische Grad, der kurz vor 2015 erhalten wurde, wie folgt:

YearX Highest_Degree 
2004 Master 
2010 PHD 
2006 Master 
NA  NA 
2004 Master 
2014 Master 

Answer 1

Ugh, was für ein schreckliches Datenformat. Wir fügen eine ID-Spalte hinzu, bereinigen sie, und dann können wir in ein paar Zeilen bekommen, was Sie wollen.

library(tidyr) 
library(dplyr) 
library(stringr) 
# create ID column 
mutate(dd, id = 1:n()) %>% 
# convert degree and year columns to long format 
gather(dd, key = "degkey", value = "degree", starts_with("Deg")) %>% 
    gather(key = "yearkey", value = "year", starts_with("Year")) %>% 
    # pull the numbers into an index 
    mutate(yr_index = str_extract(yearkey, "[0-9]+"), 
      deg_index = str_extract(degkey, "[0-9]+")) %>% 
    # get rid of junk and filter to the years you want 
    filter(yr_index == deg_index, year < 2015) %>% 
    # order by descending index 
    arrange(desc(yr_index)) %>% 
    # keep relevant columns 
    select(id, degree, year) %>% 
    # for each ID, keep the top row 
    group_by(id) %>% 
    slice(1) %>% 
    # join back to the original to complete any lost IDs 
    right_join(select(dd, id)) 

# Joining, by = "id" 
# # A tibble: 6 x 3 
# # Groups: id [?] 
#  id degree year 
# <int> <chr> <int> 
# 1  1 Master 2004 
# 2  2  PHD 2010 
# 3  3 College 2006 
# 4  4 <NA> NA 
# 5  5 Master 2004 
# 6  6 Master 2014 
# Warning message: 
# attributes are not identical across measure variables; they will be dropped 

---

diese Daten verwenden:

dd = read.table(text = "Year1 Deg_Year1 Year2 Deg_Year2 Year3 Deg_Year3 Year4 Deg_Year4 Year5 Deg_Year5 
2001 College 2004 Master  NA  NA  NA  NA  NA  NA 
2004 College 2004 Master  2010 PHD  NA  NA  NA  NA 
2006 Master  2006 College  NA  NA  NA  NA  NA  NA 
2016 Master  NA  NA   NA  NA  NA  NA  NA  NA 
2002 Master  2003 Master  2004 College 2004  Master NA  NA 
2014 Master  2017 PHD   NA  NA  NA  NA  NA  NA", 
           header = T)