How to create a database that individuals that belong to the 1st percentile of the salary variable ??

[Alexis Rodas]

Hello to all statalist users,

I have a question about how to create a base of individuals that belong to the 1st percentile or 5th percentile or etc from a total base of individuals. Let's say that to describe the percentiles, are the percentiles belonging to the salary variable.

Explaining a little more, suppose we have a base of individuals per year and quarter with the main variable: salary ("salary"). All these individuals pentenecen to any of the percentiles of salary distribution (salary). So, I want to generate a base of only the individuals that belong to the 1st percentile.

I happen to generate a dummy of the individuals that belong to the 1st percentile of salary distribution (salary). But I do not know how to create this variable, do you have any suggestions for it?

Proporciono una base de datos corta, usándola ¿podría darme una sugerencia?

* Example generated by -dataex-. To install: ssc install dataex
clear
input float(salary id)
      0   1
 385.43   2
 784.14   3
 690.34   4
 282.05   5
 466.36   6
 367.69   7
 324.61   8
1361.05   9
 374.62  10
 493.78  11
 703.49  12
 173.29  13
 156.78  14
 159.63  15
1013.57  16
1144.06  17
1598.38  18
 794.49  19
 301.05  20
 591.91  21
 178.36  22
 244.45  23
 261.33  24
  403.7  25
 749.98  26
 799.55  27
 425.66  28
1843.76  29
 934.07  30
 941.99  31
 869.07  32
 605.05  33
1107.17  34
  407.3  35
 410.82  36
1907.82  37
1236.35  38
 398.51  39
 512.37  40
 847.88  41
1587.29  42
 493.93  43
 113.86  44
 627.84  45
1061.01  46
 310.71  47
   97.2  48
     60  49
     80  50
 611.56  51
  56.93  52
 403.19  53
 525.04  54
 149.32  55
     63  56
  207.7  57
 226.46  58
1063.21  59
 766.41  60
 229.94  61
 579.08  62
  97.75  63
 368.45  64
  151.1  65
 125.48  66
 241.95  67
 156.78  68
  434.1  69
 168.74  70
  96.96  71
 132.56  72
  65.84  73
 166.84  74
 218.18  75
  62.71  76
 133.33  77
 166.66  78
  416.6  79
 167.23  80
 245.62  81
 175.66  82
 627.12  83
 281.97  84
 245.62  85
 393.06  86
 389.94  87
 308.37  88
2335.64  89
2766.72  90
1115.49  91
 684.15  92
1089.42  93
2350.87  94
2533.81  95
 722.58  96
 807.47  97
1212.98  98
 2509.3  99
2956.91 100
end

Greetings and thank you very much for the suggestions you can offer.

Alexis Rodas

[Joseph Coveney]

I happen to generate a dummy of the individuals that belong to the 1st percentile of salary distribution (salary). But I do not know how to create this variable, do you have any suggestions for it?

Try

centile salary, centile(1)
generate byte first_pctile = salary <= r(c_1) if !mi(salary)