TheLazy.Dev
Portfolio GitHub

Blog

ruby |narray |rumale |tutorial

NArray - Ruby equivalent of NumPy

kojix2

kojix2

Posted on April 25, 2019

NArray - Ruby equivalent of NumPy
ruby narray

Want to do fast calculations in Ruby? NArray is your friend!

NArray ＆ Cumo

Numo::NArray

GitHub

NArray is a powerful N-dimensional array library for science computing in Ruby. Machine learning libraries such as Rumale and Red Chainer use NArray.

Cumo

GitHub

Cumo (pronounced "koomo") is a CUDA-aware, GPU-optimized numerical library that offers a significant performance boost over Ruby Numo, while (mostly) maintaining drop-in compatibility.

NArray Data types

Integer Unsigned Integer Float Complex number
Numo::Int8 Numo::UInt8 Numo::SFloat Numo::SComplex
Numo::Int16 Numo::UInt16 Numo::DFloat Numo::DComplex
Numo::Int32 Numo::UInt32
Numo::Int64 Numo::UInt64

Import the Library 

gem install numo-narray
Enter fullscreen mode Exit fullscreen mode 
require 'numo/narray'
include Numo
Enter fullscreen mode Exit fullscreen mode

Creating Arrays 

Int32[1,2,3]
Enter fullscreen mode Exit fullscreen mode

[1, 2, 3]

Int32.new(3,3).seq
Enter fullscreen mode Exit fullscreen mode

[[0, 1, 2],
[3, 4, 5],
[6, 7, 8]]

a = [[0, 1], [2, 3]]
Int32[*a]
Enter fullscreen mode Exit fullscreen mode

[[0, 1],
[2, 3]]

Int32[1..5]
Enter fullscreen mode Exit fullscreen mode

[1, 2, 3, 4, 5]

Int32[1...5]
Enter fullscreen mode Exit fullscreen mode

[1, 2, 3, 4]

Initial Placeholders

Create an array of zeros

x = Int32.zeros(3,3)
Enter fullscreen mode Exit fullscreen mode

[[0, 0, 0],
[0, 0, 0],
[0, 0, 0]]

Create an array of ones

x = Int32.ones(3,3)
Enter fullscreen mode Exit fullscreen mode

[[1, 1, 1],
[1, 1, 1],
[1, 1, 1]]

Create an array of twos

z = y.fill 2
Enter fullscreen mode Exit fullscreen mode

[[2, 2, 2],
[2, 2, 2],
[2, 2, 2]]

Create a 3x3 identity matrix

y = Int32.eye(3,3)
Enter fullscreen mode Exit fullscreen mode

[[1, 0, 0],
[0, 1, 0],
[0, 0, 1]]

Create an array of evenly spaced values

f = DFloat.linspace(-20,20,11)
Enter fullscreen mode Exit fullscreen mode

[-20, -16, -12, -8, -4, 0, 4, 8, 12, 16, 20]

Arithmetic operations 

require 'numo/narray'
include Numo
Enter fullscreen mode Exit fullscreen mode 
a = Int32.new(3,3).seq
Enter fullscreen mode Exit fullscreen mode

[[0, 1, 2],
[3, 4, 5],
[6, 7, 8]]

b = Int32.new(3,3).seq + 1
Enter fullscreen mode Exit fullscreen mode

[[1, 2, 3],
[4, 5, 6],
[7, 8, 9]]

a + b
Enter fullscreen mode Exit fullscreen mode

[[1, 3, 5],
[7, 9, 11],
[13, 15, 17]]

b - a
Enter fullscreen mode Exit fullscreen mode

[[1, 1, 1],
[1, 1, 1],
[1, 1, 1]]

a * b
Enter fullscreen mode Exit fullscreen mode

[[0, 2, 6],
[12, 20, 30],
[42, 56, 72]]

a * 2
Enter fullscreen mode Exit fullscreen mode

[[0, 2, 4],
[6, 8, 10],
[12, 14, 16]]

a * 0.1
Enter fullscreen mode Exit fullscreen mode

[[0, 0.1, 0.2],
[0.3, 0.4, 0.5],
[0.6, 0.7, 0.8]]

Type conversion 

a = UInt8[1,2,3]
SFloat.cast(a)
# => Numo::SFloat#shape=[3]
Enter fullscreen mode Exit fullscreen mode

Dot product 

c = Int32.new(2,2).seq

# [[0, 1], 
#  [2, 3]]

d = Int32.new(2,1).seq

# [[0], 
#  [1]]

c.dot d

# [[1], 
#  [3]]
Enter fullscreen mode Exit fullscreen mode

install Numo::Linalg

Numo::Linalg.dot(c,d)
Numo::Linalg.matmul(c,d)
# [[1], 
#  [3]]
Enter fullscreen mode Exit fullscreen mode

Inspecting NArray 

e = DFloat.new(3,4).seq

# Array dimensions
e.shape              # => [3, 4]

# Length of array
e.size               # => 12

# Number of array dimensions
e.rank               # => 2
e.ndim               # => 2

# Byte size
e.byte_size          # => 96
Enter fullscreen mode Exit fullscreen mode

Convert NArray to string 

s = UInt8.ones(2,2).to_string
# => "\x01\x01\x01\x01"
# [[1, 1], 
#  [1, 1]]

UInt8.from_string(d)
# => Numo::UInt8#shape=[4]
# [1, 1, 1, 1]
Enter fullscreen mode Exit fullscreen mode

Convert NArray to Ruby Array

use to_a

a = UInt8[1,2,3]

a.to_a
Enter fullscreen mode Exit fullscreen mode

NMath

sin cos tan log, etc.

f = DFloat.linspace(-20,20,11)
# [-20, -16, -12, -8, -4, 0, 4, 8, 12, 16, 20]

NMath.sin(f)
# [-0.912945, 0.287903, 0.536573, -0.989358, 0.756802, 0, -0.756802, ...]

NMath.cos(f)
# [-0.912945, 0.287903, 0.536573, -0.989358, 0.756802, 0, -0.756802, ...]

NMath.tanh(f)
# [-1, -1, -1, -1, -0.999329, 0, 0.999329, 1, 1, 1, 1]

NMath.log(f)
# [nan, nan, nan, nan, nan, -inf, 1.38629, 2.07944, 2.48491, 2.77259, ...]

NMath.log10(f)
# [nan, nan, nan, nan, nan, -inf, 0.60206, 0.90309, 1.07918, 1.20412, ...]

NMath.sqrt(f)
# [-nan, -nan, -nan, -nan, -nan, 0, 2, 2.82843, 3.4641, 4, 4.47214]
Enter fullscreen mode Exit fullscreen mode

Subsetting, Slicing, Indexing 

a = Int32.new(10,10).seq
# [[0, 1, 2, 3, 4, 5, 6, 7, 8, 9], 
#  [10, 11, 12, 13, 14, 15, 16, 17, 18, 19], 
#  [20, 21, 22, 23, 24, 25, 26, 27, 28, 29], 
#  [30, 31, 32, 33, 34, 35, 36, 37, 38, 39], 
#  [40, 41, 42, 43, 44, 45, 46, 47, 48, 49], 
#  [50, 51, 52, 53, 54, 55, 56, 57, 58, 59], 
#  [60, 61, 62, 63, 64, 65, 66, 67, 68, 69], 
#  [70, 71, 72, 73, 74, 75, 76, 77, 78, 79], 
#  [80, 81, 82, 83, 84, 85, 86, 87, 88, 89], 
#  [90, 91, 92, 93, 94, 95, 96, 97, 98, 99]]

a[0,true]
# [0, 1, 2, 3, 4, 5, 6, 7, 8, 9]

a[true, 0]
# [0, 10, 20, 30, 40, 50, 60, 70, 80, 90]

a[1..3, 2..4]
# [[12, 13, 14], 
#  [22, 23, 24], 
#  [32, 33, 34]]

a[1]
# 1

a[-1]
# 99

a.diagonal
# [0, 11, 22, 33, 44, 55, 66, 77, 88, 99]

a.transpose
# [[0, 10, 20, 30, 40, 50, 60, 70, 80, 90], 
#  [1, 11, 21, 31, 41, 51, 61, 71, 81, 91], 
#  [2, 12, 22, 32, 42, 52, 62, 72, 82, 92], 
#  [3, 13, 23, 33, 43, 53, 63, 73, 83, 93], 
#  [4, 14, 24, 34, 44, 54, 64, 74, 84, 94], 
#  [5, 15, 25, 35, 45, 55, 65, 75, 85, 95], 
#  [6, 16, 26, 36, 46, 56, 66, 76, 86, 96], 
#  [7, 17, 27, 37, 47, 57, 67, 77, 87, 97], 
#  [8, 18, 28, 38, 48, 58, 68, 78, 88, 98], 
#  [9, 19, 29, 39, 49, 59, 69, 79, 89, 99]]

a.reverse
# [[99, 98, 97, 96, 95, 94, 93, 92, 91, 90], 
#  [89, 88, 87, 86, 85, 84, 83, 82, 81, 80], 
#  [79, 78, 77, 76, 75, 74, 73, 72, 71, 70], 
#  [69, 68, 67, 66, 65, 64, 63, 62, 61, 60], 
#  [59, 58, 57, 56, 55, 54, 53, 52, 51, 50], 
#  [49, 48, 47, 46, 45, 44, 43, 42, 41, 40], 
#  [39, 38, 37, 36, 35, 34, 33, 32, 31, 30], 
#  [29, 28, 27, 26, 25, 24, 23, 22, 21, 20], 
#  [19, 18, 17, 16, 15, 14, 13, 12, 11, 10], 
#  [9, 8, 7, 6, 5, 4, 3, 2, 1, 0]]

a.reshape(4,25)
# [[0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, ...], 
#  [25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37, 38, 39, 40, 41, 42, ...], 
#  [50, 51, 52, 53, 54, 55, 56, 57, 58, 59, 60, 61, 62, 63, 64, 65, 66, 67, ...], 
#  [75, 76, 77, 78, 79, 80, 81, 82, 83, 84, 85, 86, 87, 88, 89, 90, 91, 92, ...]

a.max
# 99

a.min
# 0

a.minmax
# [0, 99]

a.sum
# 4950

a.sum 0 
# [450, 460, 470, 480, 490, 500, 510, 520, 530, 540]

a > 49
# => Numo::Bit#shape=[10,10]
# [[0, 0, 0, 0, 0, 0, 0, 0, 0, 0], 
#  [0, 0, 0, 0, 0, 0, 0, 0, 0, 0], 
#  [0, 0, 0, 0, 0, 0, 0, 0, 0, 0], 
#  [0, 0, 0, 0, 0, 0, 0, 0, 0, 0], 
#  [0, 0, 0, 0, 0, 0, 0, 0, 0, 0], 
#  [1, 1, 1, 1, 1, 1, 1, 1, 1, 1], 
#  [1, 1, 1, 1, 1, 1, 1, 1, 1, 1], 
#  [1, 1, 1, 1, 1, 1, 1, 1, 1, 1], 
#  [1, 1, 1, 1, 1, 1, 1, 1, 1, 1], 
#  [1, 1, 1, 1, 1, 1, 1, 1, 1, 1]]

bit = (a > 49)
bit.where
# => Numo::Int32#shape=[50]
# [50, 51, 52, 53, 54, 55, 56, 57, 58, 59, 60, 61, 62, 63, 64, 65, 66, 67, ...]

a.eq 33
# [[0, 0, 0, 0, 0, 0, 0, 0, 0, 0], 
#  [0, 0, 0, 0, 0, 0, 0, 0, 0, 0], 
#  [0, 0, 0, 0, 0, 0, 0, 0, 0, 0], 
#  [0, 0, 0, 1, 0, 0, 0, 0, 0, 0], 
#  [0, 0, 0, 0, 0, 0, 0, 0, 0, 0], 
#  [0, 0, 0, 0, 0, 0, 0, 0, 0, 0], 
#  [0, 0, 0, 0, 0, 0, 0, 0, 0, 0], 
#  [0, 0, 0, 0, 0, 0, 0, 0, 0, 0], 
#  [0, 0, 0, 0, 0, 0, 0, 0, 0, 0], 
#  [0, 0, 0, 0, 0, 0, 0, 0, 0, 0]]

(a.eq 33).where
# => Numo::Int32#shape=[1]
# [33]

a[a > 90]
# [91, 92, 93, 94, 95, 96, 97, 98, 99]
Enter fullscreen mode Exit fullscreen mode 
a = Int32.new(10).seq - 5
# [-5, -4, -3, -2, -1, 0, 1, 2, 3, 4]

a.abs
# a = Int32.new(10).seq - 5
# [5, 4, 3, 2, 1, 0, 1, 2, 3, 4]

b = Int32.new(10).seq

b.cumsum
# [0, 1, 3, 6, 10, 15, 21, 28, 36, 45]
Enter fullscreen mode Exit fullscreen mode

Rnadom values 

x = DFloat.new(1000).rand
y = DFloat.new(1000).rand
Enter fullscreen mode Exit fullscreen mode

image.png

x = DFloat.new(1000).rand_norm
y = DFloat.new(1000).rand_norm
# rand_norm([mu,[sigma]])
Enter fullscreen mode Exit fullscreen mode

image.png

Statistics 

a = DFloat.new(100,100).rand_norm(1, 2)
# => Numo::DFloat#shape=[100,100]

a.size
# => 10000

a.mean
# => 0.9941970100670163

a.median
# => Numo::DFloat#shape=[]
# 1.0030267444162986

a.var
# => 3.9539947182922974

a.stddev
# => 1.9884654179271757

a.rms
# => 2.223067028599605
Enter fullscreen mode Exit fullscreen mode

Sorting Arrays 

na = Int32[1, 10, 2, 5, 9, 8, 12, 11, 3, 7, 4, 6]

na.sort

#=> Numo::Int32#shape=[12]
#[1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12]

na.max_index
# => 6

na.min_index
# => 0

na.sort_index
=> Numo::Int32#shape=[12]
[0, 2, 8, 10, 3, 11, 9, 5, 4, 1, 7, 6]
Enter fullscreen mode Exit fullscreen mode

View 

a = Int8[1,2,3,4,5]
# [1, 2, 3, 4, 5]
b = a.view
# => Numo::Int8(view)#shape=[5]
# [1, 2, 3, 4, 5]

a[0] = 0
b
# => Numo::Int8(view)#shape=[5]
# [0, 2, 3, 4, 5]

b[1] = 0
a
# => Numo::Int8#shape=[5]
# [0, 0, 3, 4, 5]
Enter fullscreen mode Exit fullscreen mode

Saving & Loading 

a = Int32.new(2,2).seq

# save
s = Marshal.dump(a)
File.binwrite("data", s)

# load
b = Marshal.load(File.read("data"))
a == b
# true
Enter fullscreen mode Exit fullscreen mode

Combining Arrays 

a = Int32[1,2,3]
b = Int32[4,5,6]

a.append b
# [1,2,3,4,5,6]

a.concatenate b
# [1,2,3,4,5,6]

a.hstack b

Int32.hstack [a,b]
# [1,2,3,4,5,6]

Int32.dstack [b,b]
# => Numo::Int32#shape=[2,3,2]
# [[[1, 1], 
#  [2, 2], 
#  [3, 3]], 
# [[4, 4], 
#  [5, 5], 
#  [6, 6]]]
Enter fullscreen mode Exit fullscreen mode

Documents

This is just some of the features of NArray. If you want to know more, please refer to these webpages.

Have a nice day!

💖 💪 🙅 🚩
kojix2
kojix2

Posted on April 25, 2019

Join Our Newsletter. No Spam, Only the good stuff.

Sign up to receive the latest update from our blog.

Related

NArray - Ruby equivalent of NumPy
ruby NArray - Ruby equivalent of NumPy

April 25, 2019