Why am I getting zeros when computing growth rates by country-year-sector in my data using tidyverse?

huangapple
go评论90阅读模式
英文:

Why am I getting zeros when computing growth rates by country-year-sector in my data using tidyverse?

问题

我想为以下数据集中的每个国家-年份-部门计算增长率：

  1. > sapply(sa1, class)
  2.      country         year       sector sector_share 
  3.     "factor"    "numeric"     "factor"    "numeric" 
  4. > print(sa1)
  5.                country year        sector sector_share
  6. 1   Sub-Saharan Africa 1981   agriculture    15.724457
  7. 2   Sub-Saharan Africa 1982   agriculture    16.165780
  8. 3   Sub-Saharan Africa 1983   agriculture    15.908671
  9. 4   Sub-Saharan Africa 1984   agriculture    17.593971
  10. 5   Sub-Saharan Africa 1985   agriculture    19.428871
  11. 6   Sub-Saharan Africa 1986   agriculture    19.593291
  12. 7   Sub-Saharan Africa 1987   agriculture    19.789807
  13. 8   Sub-Saharan Africa 1988   agriculture    20.597277
  14. 9   Sub-Saharan Africa 1989   agriculture    19.933259
  15. 10  Sub-Saharan Africa 1990   agriculture    19.790467
  17. 42  Sub-Saharan Africa 1981      industry    35.516119
  18. 43  Sub-Saharan Africa 1982      industry    32.407578
  19. ...

我使用以下代码：

  1. sa1 <- sa1 %>%
  2.   group_by(country, year, sector) %>%
  3.   arrange(year) %>%
  4.   mutate(growth_rate = ifelse(!is.na(lag(sector_share)), (sector_share / lag(sector_share) - 1) * 100, 0))

但我得到了零，这是不应该的，因为sector_share列中没有NA。

  1. > print(sa1)
  2. # A tibble: 164 × 5
  3. # Groups:   country, year, sector [164]
  4.    country             year sector        sector_share growth_rate
  5.    <fct>              <dbl> <fct>                <dbl>       <dbl>
  6.  1 Sub-Saharan Africa  1981 agriculture           15.7           0
  7.  2 Sub-Saharan Africa  1981 industry              35.5           0
  8.  3 Sub-Saharan Africa  1981 manufacturing         18.4           0
  9.  4 Sub-Saharan Africa  1981 services              44.9           0
  10.  5 Sub-Saharan Africa  1982 agriculture           16.2           0
  11.  6 Sub-Saharan Africa  1982 industry              32.4           0
  12.  7 Sub-Saharan Africa  1982 manufacturing         17.9           0
  13.  8 Sub-Saharan Africa  1982 services              46.3           0
  14.  9 Sub-Saharan Africa  1983 agriculture           15.9           0
  15. 10 Sub-Saharan Africa  1983 industry              32.3           0
  16. # ℹ 154 more rows
  17. # ℹ Use `print(n = ...)` to see more rows

我尝试计算增长率，但得到了零。这不合理，因为我的数据在sector_share列中没有NA，而且代码中我也进行了检查以防万一。

有人能帮助我吗？谢谢！

英文:

I want to compute a growth rate for each country-year-sector in the following dataset:

  1. > sapply(sa1, class)
  2.      country         year       sector sector_share 
  3.     "factor"    "numeric"     "factor"    "numeric" 
  4. > print(sa1)
  5.                country year        sector sector_share
  6. 1   Sub-Saharan Africa 1981   agriculture    15.724457
  7. 2   Sub-Saharan Africa 1982   agriculture    16.165780
  8. 3   Sub-Saharan Africa 1983   agriculture    15.908671
  9. 4   Sub-Saharan Africa 1984   agriculture    17.593971
  10. 5   Sub-Saharan Africa 1985   agriculture    19.428871
  11. 6   Sub-Saharan Africa 1986   agriculture    19.593291
  12. 7   Sub-Saharan Africa 1987   agriculture    19.789807
  13. 8   Sub-Saharan Africa 1988   agriculture    20.597277
  14. 9   Sub-Saharan Africa 1989   agriculture    19.933259
  15. 10  Sub-Saharan Africa 1990   agriculture    19.790467
  17. 42  Sub-Saharan Africa 1981      industry    35.516119
  18. 43  Sub-Saharan Africa 1982      industry    32.407578
  19. 44  Sub-Saharan Africa 1983      industry    32.303477
  20. 45  Sub-Saharan Africa 1984      industry    30.437994
  21. 46  Sub-Saharan Africa 1985      industry    30.544564
  22. 47  Sub-Saharan Africa 1986      industry    29.458658
  23. 48  Sub-Saharan Africa 1987      industry    29.490104
  24. 49  Sub-Saharan Africa 1988      industry    29.009534
  25. 50  Sub-Saharan Africa 1989      industry    29.340000
  26. 51  Sub-Saharan Africa 1990      industry    29.698078
  27. 52  Sub-Saharan Africa 1991      industry    28.727260
  29. 83  Sub-Saharan Africa 1981 manufacturing    18.419694
  30. 84  Sub-Saharan Africa 1982 manufacturing    17.895412
  31. 85  Sub-Saharan Africa 1983 manufacturing    18.037958
  32. 86  Sub-Saharan Africa 1984 manufacturing    16.316419
  33. 87  Sub-Saharan Africa 1985 manufacturing    16.256940
  34. 88  Sub-Saharan Africa 1986 manufacturing    15.728073
  35. 89  Sub-Saharan Africa 1987 manufacturing    15.825253
  36. 90  Sub-Saharan Africa 1988 manufacturing    16.320170
  37. 91  Sub-Saharan Africa 1989 manufacturing    16.062034
  38. 92  Sub-Saharan Africa 1990 manufacturing    16.134401
  39. 93  Sub-Saharan Africa 1991 manufacturing    15.826331
  41. 124 Sub-Saharan Africa 1981      services    44.946512
  42. 125 Sub-Saharan Africa 1982      services    46.323757
  43. 126 Sub-Saharan Africa 1983      services    46.071141
  44. 127 Sub-Saharan Africa 1984      services    45.820815
  45. 128 Sub-Saharan Africa 1985      services    43.226268
  46. 129 Sub-Saharan Africa 1986      services    43.409858
  47. 130 Sub-Saharan Africa 1987      services    44.298582
  48. 131 Sub-Saharan Africa 1988      services    43.191570
  49. 132 Sub-Saharan Africa 1989      services    43.023115
  50. 133 Sub-Saharan Africa 1990      services    44.043939
  51. 134 Sub-Saharan Africa 1991      services    44.995853

I use the following code:

  1. sa1 <- sa1 %>%
  2.   group_by(country, year, sector) %>%
  3.   arrange(year) %>%
  4.   mutate(growth_rate = ifelse(!is.na(lag(sector_share)), (sector_share / lag(sector_share) - 1) * 100, 0))

But I obtain zeros, which should not be since the are no NAs in the sector_share column.

  1. > print(sa1)
  2. # A tibble: 164 × 5
  3. # Groups:   country, year, sector [164]
  4.    country             year sector        sector_share growth_rate
  5.    <fct>              <dbl> <fct>                <dbl>       <dbl>
  6.  1 Sub-Saharan Africa  1981 agriculture           15.7           0
  7.  2 Sub-Saharan Africa  1981 industry              35.5           0
  8.  3 Sub-Saharan Africa  1981 manufacturing         18.4           0
  9.  4 Sub-Saharan Africa  1981 services              44.9           0
  10.  5 Sub-Saharan Africa  1982 agriculture           16.2           0
  11.  6 Sub-Saharan Africa  1982 industry              32.4           0
  12.  7 Sub-Saharan Africa  1982 manufacturing         17.9           0
  13.  8 Sub-Saharan Africa  1982 services              46.3           0
  14.  9 Sub-Saharan Africa  1983 agriculture           15.9           0
  15. 10 Sub-Saharan Africa  1983 industry              32.3           0
  16. # ℹ 154 more rows
  17. # ℹ Use `print(n = ...)` to see more rows

I tried to compute the growth rate, but I obtain zeros. It does not make sense since my data has no NAs in the sector_share column and I am doing a check even in the code just in case.

Can someone help me? Thank you!

答案1

得分: 0

由于您正在按年份分组，您的计算一次只“看到”一年，因此无法计算多年间的增长。所以不要按年份分组：

  2. library(dplyr)
  4. sa1 %>%
  5.   group_by(country, sector) %>%
  6.   arrange(year) %>%
  7.   mutate(growth_rate = ifelse(!is.na(lag(sector_share)), (sector_share / lag(sector_share) - 1) * 100, 0))
  1. # A tibble: 43 × 5
  2. # Groups:   country, sector [4]
  3.    country  year sector        sector_share growth_rate
  4.    <chr>   <int> <chr>                <dbl>       <dbl>
  5.  1 Africa   1981 agriculture           15.7       0    
  6.  2 Africa   1981 industry              35.5       0    
  7.  3 Africa   1981 manufacturing         18.4       0    
  8.  4 Africa   1981 services              44.9       0    
  9.  5 Africa   1982 agriculture           16.2       2.81 
  10.  6 Africa   1982 industry              32.4      -8.75 
  11.  7 Africa   1982 manufacturing         17.9      -2.85 
  12.  8 Africa   1982 services              46.3       3.06 
  13.  9 Africa   1983 agriculture           15.9      -1.59 
  14. 10 Africa   1983 industry              32.3      -0.321
  15. # ℹ 33 more rows
英文:

Since you’re grouping by year, your computation only “sees” one year at a time, making it impossible to compute growth across multiple years. So don’t group by year:

  2. library(dplyr)
  4. sa1 %&gt;%
  5.   group_by(country, sector) %&gt;%
  6.   arrange(year) %&gt;%
  7.   mutate(growth_rate = ifelse(!is.na(lag(sector_share)), (sector_share / lag(sector_share) - 1) * 100, 0))
  1. # A tibble: 43 &#215; 5
  2. # Groups:   country, sector [4]
  3.    country  year sector        sector_share growth_rate
  4.    &lt;chr&gt;   &lt;int&gt; &lt;chr&gt;                &lt;dbl&gt;       &lt;dbl&gt;
  5.  1 Africa   1981 agriculture           15.7       0    
  6.  2 Africa   1981 industry              35.5       0    
  7.  3 Africa   1981 manufacturing         18.4       0    
  8.  4 Africa   1981 services              44.9       0    
  9.  5 Africa   1982 agriculture           16.2       2.81 
  10.  6 Africa   1982 industry              32.4      -8.75 
  11.  7 Africa   1982 manufacturing         17.9      -2.85 
  12.  8 Africa   1982 services              46.3       3.06 
  13.  9 Africa   1983 agriculture           15.9      -1.59 
  14. 10 Africa   1983 industry              32.3      -0.321
  15. # ℹ 33 more rows

huangapple
  • 本文由 发表于 2023年5月28日 18:46:59
  • 转载请务必保留本文链接：https://go.coder-hub.com/76351083.html
  • r
  • tidyverse
匿名

发表评论

匿名网友

:?: :razz: :sad: :evil: :!: :smile: :oops: :grin: :eek: :shock: :???: :cool: :lol: :mad: :twisted: :roll: :wink: :idea: :arrow: :neutral: :cry: :mrgreen:

确定