Julia: limited printing of large arrays
As I mentioned in a comment, the way to do this in v0.5 is to use an IOContext.
A very simple way to limit the data is to pass the :limit => true parameter to the IOContext;
julia> show(IOContext(stdout, :limit => true), v)[0.147959 0.414018 … 0.595528 0.852563; 0.32679 0.824953 … 0.432143 0.036279; … ; 0.877398 0.661854 … 0.197207 0.15596; 0.0522946 0.508075 … 0.835359 0.705987]But this still does not print out the way the REPL does; that's because show with two arguments uses a single-line display. To use a multiline display, pass text/plain as the second argument to show (a MIME type):
julia> show(IOContext(stdout, :limit => true), "text/plain", v)100×100 Array{Float64,2}: 0.147959 0.414018 0.0282934 … 0.816132 0.595528 0.852563 0.32679 0.824953 0.0582351 0.822526 0.432143 0.036279 0.754989 0.724317 0.533966 0.987273 0.931932 0.973622 0.547866 0.282694 0.0295411 0.75929 0.886218 0.0813057 0.0626663 0.111795 0.625083 0.439983 0.562143 0.669046 0.712093 0.469622 0.377298 … 0.298224 0.31853 0.376066 0.774625 0.754328 0.756725 0.61113 0.76566 0.999292 0.917846 0.308363 0.489246 0.715311 0.175302 0.124059 0.310922 0.140575 0.20635 0.0280192 0.683004 0.168129 0.753361 0.755103 0.831806 0.118009 0.122374 0.281476 ⋮ ⋱ 0.420264 0.7614 0.748408 0.330983 0.0776789 0.309464 0.984379 0.851735 0.595121 0.534982 0.255317 0.743713 0.814505 0.765941 0.71852 0.730677 0.477631 0.0360992 0.910384 0.0747604 0.490685 0.0904559 0.0756424 0.313898 0.628416 0.0790874 0.401488 … 0.523521 0.397249 0.58112 0.578361 0.336352 0.261118 0.838256 0.387374 0.451647 0.66724 0.586342 0.378968 0.602694 0.450686 0.901279 0.877398 0.661854 0.685156 0.658952 0.197207 0.15596 0.0522946 0.508075 0.244423 0.95935 0.835359 0.705987 You can of course change how many rows are shown by passing in :displaysize to the IOContext:
julia> show(IOContext(stdout, :limit => true, :displaysize => (10, 10)), "text/plain", v)100×100 Array{Float64,2}: 0.147959 … 0.852563 0.32679 0.036279 0.754989 0.973622 ⋮ ⋱ 0.877398 0.15596 0.0522946 0.705987Overall, IOContext is very flexible. See its documentation for more details.
You could easily define a head function yourself. For one and two dimensions, this is pretty straightforward:
torange(n::Integer, m) = 1:min(n, m)torange(c::Colon, m) = (:)function head(a::AbstractArray{TypeVar(:T), 1}, n = 10) view(a, torange(n,size(a,1)))endfunction head(a::AbstractArray{TypeVar(:T), 2}, n1 = 10, n2 = 10) view(a, torange(n1, size(a,1)), torange(n2, size(a,2)))endThe torange method allows to use a : to return the full length in the corresponding dimension. For example
head(zeros(10, 10), 5, :)5×10 SubArray{Float64,2,Array{Float64,2},Tuple{UnitRange{Int64},Colon},false}: 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 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 0.0For more than two dimensions, I have chosen to repeat the last argument beginning from the third:
function head{Na, Nn}(a::AbstractArray{TypeVar(:T), Na}, n1 = 10, n2 = 10, ns::Vararg{TypeVar(:T), Nn} = 2) nend = last(ns) view(a, torange(n1, size(a,1)), torange(n2, size(a,2)), (torange(ns[i], size(a,i+2)) for i = 1:Nn)..., (torange(nend, size(a, i)) for i in Nn+3:Na)...)endFor example:
head(rand(10, 10, 5, 5), 3, 3, 2) # the last two is the default value and can be omitted3×3×2×2 SubArray{Float64,4,Array{Float64,4},Tuple{UnitRange{Int64},UnitRange{Int64},UnitRange{Int64},UnitRange{Int64}},false}:[:, :, 1, 1] = 0.384724 0.7328 0.585211 0.738284 0.95145 0.362914 0.43928 0.94307 0.758541[:, :, 2, 1] = 0.78603 0.588877 0.677201 0.559547 0.800559 0.488433 0.993593 0.691884 0.236595[:, :, 1, 2] = 0.25732 0.90491 0.323905 0.300924 0.703919 0.813316 0.040522 0.776142 0.624097[:, :, 2, 2] = 0.746677 0.153574 0.155539 0.991624 0.90167 0.0880094 0.0423263 0.0153597 0.0608328Note that the n-dimensional version is not type-stable, but that shouldn't matter for printing.
We can get the same output as REPL by using display() (is this what you are looking for...?) Also, the "head" and "tail" parts can be printed by using array sections, e.g.,
disp( x ) = ( display(x) ; println() ; println() )A = diagm( [ i for i=1:100 ] )disp( A )disp( A[ 1:5, : ] ) # headdisp( A[ end-4:end, : ] ) # tailB = [ i for i=1:100 ]disp( B )disp( B[ 1:5 ] ) # headdisp( B[ end-4:end ] ) # tail$ julia test.jl100x100 Array{Int64,2}: 1 0 0 0 0 0 0 0 0 0 0 0 … 0 0 0 0 0 0 0 0 0 0 2 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 3 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 4 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 5 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 6 0 0 0 0 0 0 … 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 7 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 8 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 9 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 10 0 0 0 0 0 0 0 0 0 0 0 ⋮ ⋮ ⋮ ⋱ ⋮ 0 0 0 0 0 0 0 0 0 0 0 0 92 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 93 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 94 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 95 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 … 0 0 0 0 96 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 97 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 98 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 99 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 1005x100 Array{Int64,2}: 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 2 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 3 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 4 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 5 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 05x100 Array{Int64,2}: 0 0 0 0 0 0 0 0 0 0 0 0 0 … 0 0 0 0 0 96 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 97 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 98 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 99 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 0 100100-element Array{Int64,1}: 1 2 3 4 5 6 7 8 9 10 ⋮ 92 93 94 95 96 97 98 99 1005-element Array{Int64,1}: 1 2 3 4 55-element Array{Int64,1}: 96 97 98 99 100