When initializing for example : [ [1, 2], [2, 3], [3, 2] ]
You may forget about one colon, you will get:
[ [1, 2] [2, 3], [3, 2] ]
=> [[], [3, 2]]
No syntax error:) No warnings here, beware! 
On Jul 25, 2008, at 8:58 AM, Maciej Tomaka wrote:
it’s not a syntax error, it’s the same as this
cfp:~ > ruby -e’ p( ( array = [1,2] )[ index = [2,3] ] ) ’
which is the same as
cfp:~ > ruby -e’ array = [1,2]; index = [2,3]; p array[*index]’
[]
[1, 2] [2, 3] is the array ‘[1,2]’ indexed by ‘2,3’
Ara Howard wrote:
On Jul 25, 2008, at 8:58 AM, Maciej Tomaka wrote:
it’s not a syntax error, it’s the same as this
cfp:~ > ruby -e’ p( ( array = [1,2] )[ index = [2,3] ] ) ’
which is the same as
cfp:~ > ruby -e’ array = [1,2]; index = [2,3]; p array[*index]’
[][1, 2] [2, 3] is the array ‘[1,2]’ indexed by ‘2,3’
I undestand what is going on here.
It just might be tricky to figure what’s wrong in code with such typo.
Regarding typos: Ruby can’t be expected to know what you meant to type,
so if it’s syntactically correct it will let it through – as with any
language! To do otherwise would be as annoying as Microsoft’s Clippy: "I
see you’re trying to set up an array. Did you just make a typo? "
Shadowfirebird wrote:
a = []
a << [1,2]
a << [2,3]
a << [3,4]
Without any actual understanding of what’s going on inside Ruby
(sorry!), I suspect this code will be slower, as the array has to be
grown three times, rather than springing into existence fully
initialised. Or is this a negligible overhead? Or does the code get
optimised by the compiler/interpreter? Maybe someone knowledgeable would
like to comment.
In particular it would be interesting to know if putting array pushes,
pops, shifts and unshifts inside a big loop is inefficient.
Dave
For less trivial cases I tend to:
a = []
a << [1,2]
a << [2,3]
a << [3,4]
Absolutely no chance of misreading that.
Dave B. wrote:
a << [3,4]
pops, shifts and unshifts inside a big loop is inefficient.Dave
Here’s what I got:
Rehearsal --------------------------------------
<< 0.990000 0.290000 1.280000 ( 1.331386)
[] 0.060000 0.000000 0.060000 ( 0.070595)
----------------------------- total: 1.340000sec
user system total real
<< 0.940000 0.320000 1.260000 ( 1.302148)
[] 0.060000 0.010000 0.070000 ( 0.070150)
With
ruby 1.8.6 (2008-03-03 patchlevel 114) [x86_64-linux-gnu]
and
require ‘benchmark’
include Benchmark
bmbm do |t|
t.report("<<") do
10000.times do
a = []
100.times do |x|
a << x
end
end
end
t.report("[]") do
10000.times do
a = [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]
end
end
end
-Justin
Le 29 juillet à 21:46, Justin C. a écrit :
Here’s what I got:
Rehearsal --------------------------------------
<< 0.990000 0.290000 1.280000 ( 1.331386)
[] 0.060000 0.000000 0.060000 ( 0.070595)
----------------------------- total: 1.340000secuser system total real<< 0.940000 0.320000 1.260000 ( 1.302148)
[] 0.060000 0.010000 0.070000 ( 0.070150)
Note that I tried to add this to your test :
t.report("=") do
10000.times do
a = Array.new(100)
100.times do |x|
a[x] = x
end
end
end
And…
Rehearsal --------------------------------------
<< 0.304688 0.007812 0.312500 ( 0.313559)
= 0.304688 0.000000 0.304688 ( 0.306314)
[] 0.031250 0.000000 0.031250 ( 0.027802)
----------------------------- total: 0.648438sec
user system total real
<< 0.304688 0.000000 0.304688 ( 0.309506)
= 0.296875 0.000000 0.296875 ( 0.299347)
[] 0.023438 0.007812 0.031250 ( 0.028271)
Surprising !
(ruby 1.8.6 (2007-09-24 patchlevel 111) [amd64-freebsd7])
Fred
The correct benchmark is:
(look how long x.times takes 
require ‘benchmark’
include Benchmark
bmbm do |t|
t.report("<<") do
100_000.times do
a = []
a << [ 0, 1, 2, 3, 4, 5, 6, 7, 8, 9]
a << [ 10, 11, 12, 13, 14, 15, 16, 17, 18, 19]
a << [ 20, 21, 22, 23, 24, 25, 26, 27, 28, 29]
a << [ 30, 31, 32, 33, 34, 35, 36, 37, 38, 39]
a << [ 40, 41, 42, 43, 44, 45, 46, 47, 48, 49]
a << [ 50, 51, 52, 53, 54, 55, 56, 57, 58, 59]
a << [ 60, 61, 62, 63, 64, 65, 66, 67, 68, 69]
a << [ 70, 71, 72, 73, 74, 75, 76, 77, 78, 79]
a << [ 80, 81, 82, 83, 84, 85, 86, 87, 88, 89]
a << [ 90, 91, 92, 93, 94, 95, 96, 97, 98, 99]
end
end
t.report("[]") do
100_000.times do
a = [
[ 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]
]
end
end
t.report("NO") do
100_000.times do
100.times do |x|
end
end
end
end
Rehearsal --------------------------------------
<< 1.020000 0.020000 1.040000 ( 1.047729)
[] 0.730000 0.010000 0.740000 ( 0.741970)
NO 4.270000 1.510000 5.780000 ( 5.808074)
----------------------------- total: 7.560000sec
user system total real
<< 1.000000 0.020000 1.020000 ( 1.032165)
[] 0.740000 0.010000 0.750000 ( 0.743470)
NO 5.070000 1.560000 6.630000 ( 6.634026)
So << and [] are comparable.
Shadowfirebird has good solution inmho w/o significant perfomance
impact.
tested again on 1.8.6,
previous was on 1.8.4 (2005-12-24) [i486-linux] 1.8.4)
Rehearsal --------------------------------------
<< 0.742188 0.000000 0.742188 ( 0.748359)
[] 0.570312 0.000000 0.570312 ( 0.567802)
NO 0.007812 0.000000 0.007812 ( 0.009996)
----------------------------- total: 1.320312sec
user system total real
<< 0.742188 0.000000 0.742188 ( 0.747654)
[] 0.570312 0.000000 0.570312 ( 0.567690)
NO 0.007812 0.000000 0.007812 ( 0.009741)
ruby 1.8.6 (2007-09-24 patchlevel 111) [i386-freebsd6]
Maciej Tomaka wrote:
tested again on 1.8.6,
previous was on 1.8.4 (2005-12-24) [i486-linux] 1.8.4)Rehearsal --------------------------------------
<< 0.742188 0.000000 0.742188 ( 0.748359)
[] 0.570312 0.000000 0.570312 ( 0.567802)
NO 0.007812 0.000000 0.007812 ( 0.009996)
----------------------------- total: 1.320312secuser system total real<< 0.742188 0.000000 0.742188 ( 0.747654)
[] 0.570312 0.000000 0.570312 ( 0.567690)
NO 0.007812 0.000000 0.007812 ( 0.009741)ruby 1.8.6 (2007-09-24 patchlevel 111) [i386-freebsd6]
The ‘NO’ test was wrong.
This is retest with code that was pasted before:
Rehearsal --------------------------------------
<< 0.750000 0.000000 0.750000 ( 0.749043)
[] 0.570312 0.000000 0.570312 ( 0.571285)
NO 1.054688 0.000000 1.054688 ( 1.056768)
----------------------------- total: 2.375000sec
user system total real
<< 0.750000 0.000000 0.750000 ( 0.745733)
[] 0.562500 0.000000 0.562500 ( 0.567207)
NO 1.039062 0.000000 1.039062 ( 1.040605)
Le 30 juillet à 11:24, Maciej Tomaka a écrit :
The correct benchmark is:
/…/
In both cases, your arrays are nested ; you’re not testing the same
thing…
Fred
F. Senault wrote:
Le 30 juillet � 11:24, Maciej Tomaka a �crit :
The correct benchmark is:
/…/
In both cases, your arrays are nested ; you’re not testing the same
thing…Fred
Please re-read the posts. We were measuring difference between
initializations:
a = [ [1,2,3], [2,3,4], [5,6,7] ]
and
a = []
a << [1,2,3]
a << [2,3,4]
a << [5,6,7]
And that is what I was testing.
Best Regards.
Maciej