forked from denizyuret/Knet.jl
-
Notifications
You must be signed in to change notification settings - Fork 0
/
julia-tutorial.jl
279 lines (252 loc) · 5.37 KB
/
julia-tutorial.jl
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
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
# Julia tutorial. Concepts:
# - Pkg: clone, add, rm, update, init, status, using
# - Arrays: ranges, tuples, multidimensional arrays, indexing, eltype, size, length, sizeof, ndims, transpose, mean, std, minimum, maximum, reshape, vec, randn, rand, zeros, ones, unary ops, element-wise ops, broadcasting ops, reductions, sub, map, copy, copy!, == vs ===, @time, CudaArray, concatenation
# - Functions: if, for, function, anonymous function, map, optional args, keyword args, three-dot notation, Dict, Symbol
# Packages
julia> Pkg.init()
julia> Pkg.add("Knet")
julia> Pkg.status()
julia> Pkg.add("ASCIIPlots")
julia> using ASCIIPlots
julia> a = eye(3)
julia> imagesc(a)
julia> Pkg.rm("ASCIIPlots")
julia> Pkg.update()
# Arrays: indexing
julia> a=reshape([1:15;], (3,5))
julia> a[1,1]
julia> a[3,5]
julia> a[4]
# Ranges
julia> 1:10
julia> for i=1:10; println(i); end
julia> for i=1:3:10; println(i); end
julia> 1:100000000000
julia> [1:10;]
julia> a=1:10
julia> typeof(a)
julia> a=1:0.5:10
julia> typeof(a)
julia> [a;]
julia> collect(a)
# Indexing continued
julia> a[3,1:2]
julia> a[2:3,1:2]
julia> a
julia> a[2:3,:]
julia> b=[1,3,5]
julia> a[:,b]
julia> b=[false, true, false, true, false]
julia> a[:,b]
julia> a
julia> a[10]
julia> a
julia> a[1]
julia> a[3:5]
julia> a = randn(10)
julia> a[a.>0]
julia> a.>0
julia> a
julia> b = (a.>0)
julia> a[b]
# Arrays: basic ops
julia> a = zeros(3,5)
julia> eltype(a)
julia> size(a)
julia> size(a,1)
julia> size(a,2)
julia> length(a)
julia> sizeof(a)
julia> ndims(a)
julia> transpose(a)
julia> a.'
julia> a'
# Arrays: unary ops
julia> a = rand(3,5)
julia> sin(a)
julia> -a
julia> 10a
# Arrays: binary ops
julia> a = rand(3,5)
julia> b = rand(3,5)
julia> a + b
julia> a * b => matmul
julia> a .* b => element-wise
julia> a
julia> b
julia> a .+ b
julia> a + b
julia> b = rand(5,2)
julia> a * b
julia> a = rand(2000,1000)
julia> b = rand(1000, 3000)
julia> c = a*b
julia> a = rand(10000, 2000)
julia> b = rand(2000, 5000)
julia> c = a * b
julia> map(size, (a,b,c))
# Arrays: broadcasting
julia> a = rand(3,5)
julia> a > 0.5
julia> a .> 0.5
julia> b = rand(3,1)+10
julia> a + b
julia> a .+ b
julia> a
julia> b
julia> broadcast(+, a, b)
julia> a + b
# Arrays: reductions
julia> a = rand(3,5)
julia> mean(a)
julia> maximum(a)
julia> minimum(a)
julia> mean(a)
julia> mean(a,1)
julia> mean(a,2)
julia> minimum(a,1)
# Arrays: slices vs subarrays
julia> a = rand(3,5)
julia> b = a[1:2,1:2]
julia> b[1,1]=0
julia> b
julia> a
julia> b = sub(a, 1:2, 1:2)
julia> a
julia> b
julia> b[1,1]=0
julia> b
julia> a
# Arrays: concatenation
julia> a = rand(3,3)
julia> b = rand(3)
julia> hcat(a,b)
julia> [a b]
julia> vcat(a,b) => Error
julia> vcat(a,b')
julia> [a; b']
# Arrays: reshaping
julia> a = reshape([1:15.0;], 3, 5)
julia> vec(a)
julia> a[1,2]
julia> a[2,2]
julia> a[2,3]
julia> a[2,4]
julia> a
julia> reshape(a, (5,3))
julia> a
julia> size(a,1)
julia> a
# Arrays: copying and identity
julia> a
julia> d = copy(a)
julia> d[1]=0
julia> d
julia> a
julia> a===d
julia> copy!(d, a)
julia> d
julia> a
julia> a==d
julia> a===d
julia> f = a
julia> a==f
julia> a===f
julia> a==d
julia> a===d
# KnetArrays: (for GPU machines)
julia> a = rand(10000, 2000)
julia> b = rand(2000, 5000)
julia> @time c = a * b;
julia> @time c = a * b;
julia> @time c = a * b;
julia> @time c = a * b;
julia> using Knet
julia> a2 = KnetArray(a)
julia> b2 = KnetArray(b)
julia> @time c2 = a2 * b2;
julia> c3 = Array(c2)
julia> isapprox(c,c3)
# Conditionals and Loops:
julia> if 1 > 0; println(:foo); else; println(:bar); end
julia> if 1 > 0; println(:foo); elseif 2 > 3; println(:bar); else 0
julia> for i=1:10; println(i); end
# Functions:
julia> function f(x,y); x+y; end
julia> f(2,3)
# Functions and files:
julia> pwd()
julia> cd("tutorial")
julia> include("foo.jl")
julia> foo(4,5)
julia> include("foo.jl")
julia> foo(4,5)
julia> include("foo.jl")
julia> foo(4,5)
julia> include("foo.jl")
julia> foo(4,5)
# Tuples
julia> a=(1,2,3)
julia> typeof(a)
julia> a[1]
julia> a[1] = 10 => Error
# Functions that return tuples:
julia> include("foo.jl")
julia> foo(4,5) => (23,9)
julia> var1,var2 = foo(4,5)
julia> var1
julia> var2
# Functions: anonymous
julia> function (x); x+1; end
julia> x->x+1
julia> a = rand(10)
julia> map(x->2x, a)
julia> bar(x,y)=2x+3y
julia> bar(3,5)
julia> foo(3,5)
# Functions: first class entities
julia> baz = foo
julia> baz(3,5)
julia> baz === foo
julia> baz == foo
# Functions: optional arguments
julia> include("foo.jl")
julia> foo(3,5) => (21,8)
julia> foo(3) => (15,6)
julia> foo() => Error
julia> foo(1,2,3) => Error
# Functions: keyword arguments
julia> include("foo.jl")
julia> foo(1,2)
julia> foo(1,2,scale=100)
julia> foo(1,2;scale=100)
julia> foo(1,2,100) => Error
julia> include("foo.jl")
julia> foo(1,2)
julia> foo(1,2,scale=100)
julia> foo(1,2,bias=30)
julia> foo(1,2,bias=30,scale=200)
# Functions variable number of arguments
julia> f(x,y...)=println("x=$x y=$y")
julia> f()
julia> f(3)
julia> f(3,4)
julia> z(x,y...)=println("x=$x y=$y")
julia> z(3,4)
julia> z(3,4,5,6,7,8)
julia> q(x,y; scale=10, kwargs...)=println("x=$x y=$y scale=$scale kwargs=$kwargs")
julia> q(1,2)
julia> q(1,2; scale=3)
julia> q(1,2; fish=6)
julia> q(1,2; fish=6, dog=1)
# Dictionaries
julia> d=Dict()
julia> d[:foo]=1
julia> d[:bar]=2
julia> d
julia> d[:bar]
julia> d[:baz]
# Dictionaries as keyword argument lists
julia> q(1,2; d...)
julia> q2(x,y; scale=10, kwargs...)=println("x=$x y=$y scale=$scale kwargs[foo]=$(Dict(kwargs)[:foo])")
julia> q2(1,2; d...)