I tried wrapping your code in a function and compiling it by running it
once as a warmup. That does help, I went from 12s to 7s after warmup. But
you're still really far from C++. I don't really know anything about
DataFrame, but I suspect your problem is with the generating the
subsets(the
More generally, is there anything like futures/callbacks or async IO in
Julia? I couldn't find anything, but maybe I just don't know the right
keywords.
On Sunday, August 23, 2015 at 1:01:28 AM UTC+2, Michael Wang wrote:
I am new to Julia. I heard that Julia has the performance with Cpp even
though it is a high level language. I tested an example on my machine,
however, the result was that Julia was in the same ballpark with R not with
If the type hierarchy is implemented well, functions for `Dict` would
ideally be written to work with `super(Dict)` (i.e., `Associative`). And
in fact, some are--e.g., `in`, `haskey`, `show`, `keys`, `values`, etc.
See `base/dict.jl` for the implementation.
Now, it could be argued that the type
Thanks for the detailed explanation!
Actually, I did not set up my github ssh keys yet, but I am using HTTPS.
Uwe
Am Sonntag, 23. August 2015 18:17:10 UTC+2 schrieb Timothée Poisot:
If you use 2-factors authentication, it seems that you have to push
METADATA manually.
What's good with Julia is that if something is if a library or something is
slow it is usually easy to tweak it to get your performance.
For example, if we only focus on the beta regressions, here is your current
code (a little modified to remove the benchmarking from reading the csv):
What's good with Julia is that if something is slow then you can usually
easily re
For example, if we only focus on the beta regressions, here your current
code (a little modified to remove the benchmarking from reading the
csv): https://gist.github.com/KristofferC/f424e1d7daf4c6496ad2
On my
There's still this related problem somewhere in the Docs.
See: the maxabs at http://docs.julialang.org/en/latest/stdlib/collections/
On Wednesday, August 12, 2015 at 1:57:15 PM UTC+2, Yichao Yu wrote:
On Wed, Aug 12, 2015 at 7:42 AM, J Luis jmf...@gmail.com javascript:
wrote:
In the
If you use 2-factors authentication, it seems that you have to push
METADATA manually. https://github.com/JuliaLang/julia/issues/8228 seems
to say it's possible, but I never managed to get it to work.
Le samedi 22 août 2015 à 22:26 -0700, Tony Kelman a écrit :
Most likely your github ssh keys
Le dimanche 23 août 2015 à 09:22 -0700, Andrew a écrit :
I tried wrapping your code in a function and compiling it by running
it once as a warmup. That does help, I went from 12s to 7s after
warmup. But you're still really far from C++. I don't really know
anything about DataFrame, but I
You can use Pkg.publish() with 2FA. It's a bit tricky -
see https://github.com/JuliaLang/julia/issues/10766#issuecomment-90788604
for a possible workaround.
On Sunday, August 23, 2015 at 9:17:10 AM UTC-7, Timothée Poisot wrote:
If you use 2-factors authentication, it seems that you have to
Hello all,
Apologies in advance if I’m missing something obvious. I’ve only just started
experimenting with some of the more advanced features of Julia’s types, and
I’ve hit a bit of a wall…
Say I wanted to replicate something like Python’s `collections.defaultdict` or
Ruby’s `Hash` class
Which indeed seems to be within 2x c++ speed. I don't think it's fair to
use DFs in R and Julia if they're not used in c++. I wonder what could be
done to the R implementation.
On Sunday, August 23, 2015 at 9:46:04 PM UTC+2, Kristoffer Carlsson wrote:
What's good with Julia is that if
Calling error(message) should report an error from the function in which
it is called, and it would be helpful to show the call chain. Currently,
calling error reports an error from the top of the call chain, the function
generating the error is not shown. Throwing an exception is either
On Sunday, August 23, 2015 at 1:47:06 PM UTC+2, Sisyphuss wrote:
I hope this package can be based on the data structure `AbstractArray`
instead of `DataFrame`.
Can you provide some background why and where's the problem with DataFrame
in which plotting package?
I'm not sure it'll have *that* impact but thanks anyway! The latest version
also shows the Julia version it applies to. That should at least help.
Thank you Curtis.
On Saturday, August 22, 2015 at 2:03:20 PM UTC-4, Eric Davies wrote:
Here they are! https://github.com/quinnj/TimeZones.jl
IMO it's better than pytz, which is exciting.
All credit goes to Curtis Vogt @omus.
I am not expert of the DataFrame package. So I cannot tell the downside of
DataFrame. However, since DataFrame is a part of the standard library, in
my opinion, the support for build-in structures such as AbstractArray
should be the priority.
As I know, most functionalities of Gadfly is
I hope this package can be based on the data structure `AbstractArray`
instead of `DataFrame`.
On Sunday, August 23, 2015 at 6:34:38 AM UTC+2, wil...@gmail.com wrote:
Hi all,
I've started working on a Julia PLplot wrapper
https://github.com/wildart/PLplot.jl. PLplot is a powerful
Scott makes a good point. Thank you Jacob.
On Sun, Aug 23, 2015 at 6:33 AM, Scott Jones scott.paul.jo...@gmail.com
wrote:
I like ask the things that Jacob is working on for Julia, really important
stuff for general programming, this, strings, databases, CSV, etc.
+100
Yes, this is a common way to do this. Usually to get the non-mutating behavior
you need to make a copy first anyway. (Some languages just do so implicitly and
don't allow anything else.) Sometimes, as Kristoffer mentions, there might be a
way to construct a new object directly instead of
Firstly read the following
thoroughly: http://docs.julialang.org/en/release-0.3/manual/performance-tips/
Secondly I'd try to avoid IO when benchmarking unless IO is a necessary
part of the operation.
My guess is you'll get a big speedup just by putting everything in a
function, but there'll
Potentially. It's worth experimenting with different approaches, profiling
as you iterate with the code to get a feel for what are going to be
idiomatic and high-performance methods of solving problems in Julia. Trying
to port code directly from any significantly different language is not
I like ask the things that Jacob is working on for Julia, really important
stuff for general programming, this, strings, databases, CSV, etc.
+100
It will probably double the rate of new Julia adopters! This should go to
the website. Great job and idea.
Em segunda-feira, 17 de agosto de 2015 14:07:02 UTC-3, Ian Hellström
escreveu:
Hi everyone,
I just created a cheat page for Julia: http://tinyurl.com/JuliaLang.
It's full of
Ah, it will be fixed by https://github.com/JuliaLang/julia/pull/12760
On Sunday, August 23, 2015 at 2:46:24 PM UTC+2, Sisyphuss wrote:
There's still this related problem somewhere in the Docs.
See: the maxabs at http://docs.julialang.org/en/latest/stdlib/collections/
On Wednesday, August
It is of course possible to write slow julia code. Most of the time in your
code is spent in the @where macro in DataFramesMeta so you should look
there to see if things can be improved.
On Sunday, August 23, 2015 at 7:30:59 AM UTC+2, Cedric St-Jean wrote:
Also, Julia compiles code on the
Thanks for the pointers!
Out of curiosity, what impact does parameterizing the promotion rule have? I
tried it both ways but couldn’t detect any difference experimenting with
`promote`.
On August 24, 2015 at 00:11:11, Sisyphuss (zhengwend...@gmail.com) wrote:
IMHO, there are two things
On August 23, 2015 at 21:08:16, Kevin Squire (kevin.squ...@gmail.com) wrote:
If the type hierarchy is implemented well, functions for `Dict` would ideally
be written to work with `super(Dict)` (i.e., `Associative`). And in fact, some
are--e.g., `in`, `haskey`, `show`, `keys`, `values`, etc.
IMHO, there are two things missing:
1) You should define outer constructor for `DefaultDict(val, Dict())`;
2) You should make your promotion rules parametric.
On Sunday, August 23, 2015 at 7:48:11 PM UTC+2, Joshua Ballanco wrote:
Hello all,
Apologies in advance if I’m missing something
nvm, dumb error on my part
On Sunday, August 23, 2015 at 4:45:15 PM UTC-4, Edward Chen wrote:
Would there be any reason to suspect that the ordering of the eigenvalues
and the eigenvectors are different? I am doing a side-by-side comparison
with the eig function in MATLAB and it seems to
Very silly question. In Julia 0.3, I noticed that after slicing some
DataFrame, the resulting object doesn't keep its original indexes. For
example:
julia using DataFrames
julia df = DataFrame(x=1:100)
julia df[20:50,:]
31x1 DataFrame
| Row | x |
|-||
| 1 | 20 |
| 2 | 21 |
| 3
Update: I renamed the page to The Fast Track to Julia so as not to offend
anyone. There were some people who questioned the double entendre, so I
thought it best to remove it completely.
On Sunday, 23 August 2015 14:39:43 UTC+2, Ian Hellström wrote:
I'm not sure it'll have *that* impact but
On Sunday, August 23, 2015 at 10:18:23 PM UTC+2, Patrick Kofod Mogensen
wrote:
Which indeed seems to be within 2x c++ speed. I don't think it's fair to
use DFs in R and Julia if they're not used in c++. I wonder what could be
done to the R implementation
If it something commonly
Hi Steven,
thanks for your answer! It turns out I misunderstood @async long time ago,
assuming it also makes a remote call to other processes and thus introduces
true multi-tasking. So now I need to rethink my approach before going
further.
Just to clarify: my goal is to perform as many
Would there be any reason to suspect that the ordering of the eigenvalues
and the eigenvectors are different? I am doing a side-by-side comparison
with the eig function in MATLAB and it seems to give a different ordering
of vectors for the same matrix.
On Saturday, April 26, 2014 at 5:15:21 PM
Is Juno only an editor (I should try, but I haven't..) as you say. They
say, environment. Should you say IDE?
If/when Julia adoption doubles, wee never know - maybe the cheat sheet
helped a lot..:) To me Julia is good enough (an understatement) and people
just have to know about it and have the
@parallel in Julia is for executing separate parallel processes (true
multi-tasking, with separate address spaces). @async is for tasks, which
are green threads and represent cooperative multitasking (within the same
process and the same address space).
I/O in Julia is asynchronous — while
DataFrames don't have real row indexes. If you want a row identifier, you
can add it as an extra column.
On Aug 23, 2015 4:32 PM, Robert Smith rsmith31...@gmail.com wrote:
Very silly question. In Julia 0.3, I noticed that after slicing some
DataFrame, the resulting object doesn't keep its
to replace with something that cannot be caught, but for this one I cannot
suggest a use case (maybe writing way past array boundaries when using
@inbounds?)
Maybe a better example for something that should not be possible to catch
is an assertion error.
On Sat, Aug 22, 2015 at 5:06 PM,
Unfortunately, when it comes to displaying the columns in a front-end,
there are actually two real columns. For example, this is the output using
IJulia:
In [18]: kc.get_iopub_msg(timeout=1)
Out[18]:
{'buffers': [],
'content': {u'data': {u'text/html': u'table
On Monday, August 24, 2015 at 12:38:10 PM UTC+10, Michele Zaffalon wrote:
to replace with something that cannot be caught, but for this one I
cannot suggest a use case (maybe writing way past array boundaries when
using @inbounds?)
Maybe a better example for something that should
to replace with something that cannot be caught, but for this one I cannot
suggest a use case (maybe writing way past array boundaries when using
@inbounds?)
Maybe a better example for something that should not be possible to catch
is an assertion error.
But then catching assertions is
Hi all,
We've recently reworked the handling of Colors by Julia into a few new
packages:
https://github.com/JuliaGraphics/ColorTypes.jl
https://github.com/JuliaGraphics/Colors.jl
https://github.com/JuliaGraphics/ColorVectorSpace.jl
and are in the process of switching most of the julia package
Scott, I think you are misunderstanding. In this case the fancy macros on
the DataFrames slows things down instead of making it faster. If the C++
code was allowed to much with the raw data then it should be fair to allow
the Julia code to also do that for a performance increase. Without the
Ah, thanks! I thought using DF was considered the unfair advantage!
Table driven implementation, so the logic is easily changed to suit.
Introduces (afaik) two operators that push Maybe to True, False
respectively.
https://github.com/J-Sarnoff/ThreeValued.jl
On Sunday, August 23, 2015 at 10:14:11 PM UTC-4, Jeffrey Sarnoff wrote:
Table driven implementation, so the logic is easily changed to suit.
Introduces (afaik) two operators that push Maybe to True, False
respectively.
48 matches
Mail list logo