Question

我有一个看起来像这样的表：

ID    Start_year    Status_2005    Status_2006    Status_2007
 1          2005            GBR            GBR            FRA
 2          2006             NA            FRA            FRA
 3          2007             NA             NA            GBR
 4          2006             NA            UKR            RUS

我想对数据进行整形，以便给出起始年份之后年份的状态。所以上面看起来像这样：

ID     Year_0    Year_1    Year_2
 1        GBR       GBR       GBR
 2        FRA       FRA        NA
 3        GBR        NA        NA
 4        UKR       RUS        NA

我一直在尝试在R中使用tidyverse，结合使用collect和“ starts_with”，并进行变异以创建新列。但是，我始终以“ years_since_start_year”这一列作为结尾，并且无法弄清楚如何扩展此列以构成最终表。

非常感谢任何帮助

Answer 1

这是data.table方法：

exports.obter_lugares_zona = (req, res, next) => {
Zona.find({_id: req.params.id_area})
    .then(data => {
            let flat = data.reduce((data, {description, streets}) => {
                streets.forEach(({name, places}) => {
                    places.forEach(({loc, spotType, used}) => {
                        if (req.query.type === "Handicap") {
                            data.push(loc, spotType, used)
                        }
                    })
                })
                return data
            }, []);
            res.json(flat);
        }
    )
    .catch(error => {
        return next(error);
    });
}

Answer 2

这就是我使用tidyverse的方式

library(tidyverse)

# create data
df_raw <- data.frame(ID = c(1:4), 
                 Start_year = c(2005,2006,2007,2006),
                 Status_2005 =c("GBR", NA, NA, NA),
                 Status_2006 =c("GBR", "FRA", NA, "UKR"),
                 Status_2007 =c("FRA", "FRA", "GBR", "RUS"),
                 stringsAsFactors = F)



df <- df_raw %>% 
  gather(starts_with("Status"), key = Key, value = Value ) %>% 
  arrange(ID) %>% 
  drop_na(Value) %>% 
  mutate(cnt = unlist(map(rle(ID)$lengths-1, seq, from = 0, by =1 ))) %>% 
  mutate(Key = paste0("Year_", cnt)) %>% 
  select(-Start_year, -cnt) %>% 
  spread(key = Key, value = Value)

df
#>   ID Year_0 Year_1 Year_2
#> 1  1    GBR    GBR    FRA
#> 2  2    FRA    FRA   <NA>
#> 3  3    GBR   <NA>   <NA>
#> 4  4    UKR    RUS   <NA>

Answer 3

这里有一个粗略的基础R + dplyr：

df %>%
  select(starts_with("Status")) %>%
  apply(1, function(x) {x <- x[!is.na(x)]; length(x) <- 3; x}) %>%
  t() %>%
  as.data.frame() %>%
  cbind(df[["ID"]], .) %>%
  setNames(c("ID", paste0("Year_", 1:3)))

  ID Year_1 Year_2 Year_3
1  1    GBR    GBR    FRA
2  2    FRA    FRA   <NA>
3  3    GBR   <NA>   <NA>
4  4    UKR    RUS   <NA>

Tidyverse样式：

library(tidyr)
library(dplyr)
df %>%
  select(-Start_year) %>%
  gather(key = "year", value = "country", -ID) %>%
  filter(!is.na(country)) %>%
  group_by(ID) %>%
  mutate(year = paste0("year_", 1:length(year))) %>%
  spread(key = "year", value = "country")

# A tibble: 4 x 4
# Groups:   ID [4]
     ID year_1 year_2 year_3
  <int> <chr>  <chr>  <chr> 
1     1 GBR    GBR    FRA   
2     2 FRA    FRA    NA    
3     3 GBR    NA     NA    
4     4 UKR    RUS    NA

根据另一列中的值重塑状态列

3 个答案: