Percentiles de cálculo (rubí)
-
25-09-2019 - |
Pregunta
Mi código se basa en los métodos descritos aquí y aquí .
def fraction?(number)
number - number.truncate
end
def percentile(param_array, percentage)
another_array = param_array.to_a.sort
r = percentage.to_f * (param_array.size.to_f - 1) + 1
if r <= 1 then return another_array[0]
elsif r >= another_array.size then return another_array[another_array.size - 1]
end
ir = r.truncate
another_array[ir] + fraction?((another_array[ir].to_f - another_array[ir - 1].to_f).abs)
end
Ejemplo de uso:
test_array = [95.1772, 95.1567, 95.1937, 95.1959, 95.1442, 95.061, 95.1591, 95.1195,
95.1065, 95.0925, 95.199, 95.1682]
test_values = [0.0, 0.1, 0.2, 0.3, 0.4, 0.5, 0.6, 0.7, 0.8, 0.9, 1.0]
test_values.each do |value|
puts value.to_s + ": " + percentile(test_array, value).to_s
end
Salida:
0.0: 95.061
0.1: 95.1205
0.2: 95.1325
0.3: 95.1689
0.4: 95.1692
0.5: 95.1615
0.6: 95.1773
0.7: 95.1862
0.8: 95.2102
0.9: 95.1981
1.0: 95.199
El problema aquí es que el percentil 80 es mayor que el 90º y el 100º. Sin embargo, por lo que yo puedo decir a mi aplicación es como se ha descrito, y devuelve la respuesta correcta para el ejemplo dado (0,9).
¿Hay un error en mi código que no estoy viendo? ¿O hay una mejor manera de hacer esto?
Solución
guión
Esto suena como un problema de tarea. De todos modos, que era un poco divertido de hacer.
# Score class
class Score
attr_accessor :value, :percentile
def initialize(score)
self.value = score.to_f
end
def <=>(foo)
self.value <=> foo.value
end
end
# load scores
scores = []
DATA.each do |line|
scores << Score.new(line)
end
scores.sort!
scores_count = scores.size
# iterate through scores and calculate percentile
scores.each_with_index do |s, i|
# L/N(100) = P
# L = number of scores beneath this score (score array index)
# N = total number of scores
# P = percentile
s.percentile = (i.to_f/scores_count.to_f*100).ceil
end
# output
puts "What is the precise percentile of each score"
scores.each_with_index do |s,i|
puts "#{s.value} is in the #{s.percentile} percentile"
end
# bonus: what score is in the Xth percentile?
puts "\nWhat score is in the Xth percentile?"
percentiles = [0, 10, 20, 30, 40, 50, 60, 70, 80, 90, 100]
percentiles.each do |p|
# P/100(N) = L
# P = percentile
# N = total number of scores
# L = score array index
l = (p.to_f/100*scores_count).ceil
puts "#{p} percentile? #{scores[l].value}"
end
__END__
95.1772
95.1567
95.1937
95.1959
95.1442
95.061
95.1591
95.1195
95.1065
95.0925
95.199
95.1682
salida
What is the precise percentile of each score
95.061 is in the 0 percentile
95.0925 is in the 9 percentile
95.1065 is in the 17 percentile
95.1195 is in the 25 percentile
95.1442 is in the 34 percentile
95.1567 is in the 42 percentile
95.1591 is in the 50 percentile
95.1682 is in the 59 percentile
95.1772 is in the 67 percentile
95.1937 is in the 75 percentile
95.1959 is in the 84 percentile
95.199 is in the 92 percentile
What score is in the Xth percentile?
0 percentile? 95.061
10 percentile? 95.1065
20 percentile? 95.1195
30 percentile? 95.1442
40 percentile? 95.1567
50 percentile? 95.1591
60 percentile? 95.1772
70 percentile? 95.1937
80 percentile? 95.1959
90 percentile? 95.199
Otros consejos
tengo trabajo. Añadido -Infinity
a la matriz de modo que pudiera utilizar los índices en el rango 1 - N
. También me multiplicando el valor en la última línea de la variable equivocada.
def percentile(param_array, percentage)
another_array = param_array.to_a.dup
another_array.push(-1.0/0.0) # add -Infinity to be 0th index
another_array.sort!
another_array_size = another_array.size - 1 # disregard -Infinity
r = percentage.to_f * (another_array_size - 1) + 1
if r <= 1 then return another_array[1]
elsif r >= another_array_size then return another_array[another_array_size]
end
ir = r.truncate
fr = fraction? r
another_array[ir] + fr*(another_array[ir+1] - another_array[ir])
end
La línea r = ...
puede ser reemplazado por r = percentage.to_f * (another_array_size + 1)
utilizar la fórmula en el primer enlace en lugar de Excel de.
Salida:
0.0: 95.061
0.1: 95.0939
0.2: 95.1091
0.3: 95.12691
0.4: 95.1492
0.5: 95.1579
0.6: 95.16456
0.7: 95.1745
0.8: 95.1904
0.9: 95.19568
1.0: 95.199
También puede monkeypatch Enumerable:
module Enumerable
def rank value, n_tiles
count = self.length
raise "You cannot split an array of #{count} elements into #{n_tiles} tiles!" if n_tiles > count
ordered_array = self.sort
split_size = count / n_tiles
boundaries = []
(n_tiles - 1).times do |i|
boundaries << ordered_array[(i + 1) * split_size - 1]
end
boundaries.each_with_index do |boundary, i|
if value > boundaries.last
return n_tiles
elsif value <= boundary
return (i + 1)
end
end
end
end
Después de esto usted sería capaz de hacer algo como esto:
a = [1,4,2,5,3,6]
# Test in which range (rank) the number '1' would be places, if the array is ordered and spit into 3 pieces:
a.rank(1,3)
#=> 1