Re: mat file opening

[Levente Torok]

Hi Guys,

I was stupid. I did a fromdos conversion for all files in the package
which ate two \r characters from the data files. This is it.
Lev

On Sun, Apr 15, 2012 at 10:16 PM, Levente Torok <address@hidden> wrote:
> Thanks Martin for checking.
>
> In this case my version of octave was the key of the problem.
> WeIl I didn't mean to compile from sources but I will need to on my ubuntu.
>
> Lev
>
> On Sat, Apr 14, 2012 at 10:36 PM, Martin Helm <address@hidden> wrote:
>> Am 14.04.2012 22:26, schrieb Martin Helm:
>>> Am 14.04.2012 21:59, schrieb Levente Torok:
>>>> HI Guys,
>>>>
>>>> I checked iagain nips_wd.dat doesn't open in my octave 3.2.4.
>>>> I asked the author to reproduce cora_wd.dat file in order to make sure
>>>> that this is not a transmission error kind of thing.
>>>>
>>>> I have no access to matlab so I have no clue how matlab would behave
>>>> but Ben tested it.
>>>>
>>>> Thanks,
>>>> Lev
>>>>
>>> Maybe a problem with the 3.2.4 version, the nips_wd seems so far to load
>>> fine on 3.6.1
>>>
>>> octave:1> load nips_wd.mat
>>> octave:2> whos
>>> Variables in the current scope:
>>>
>>>    Attr Name         Size                     Bytes  Class
>>>    ==== ====         ====                     =====  =====
>>>         nips_wd  13649x1740                11206504  double
>>>
>>> Total is 933295 elements using 11206504 bytes
>>>
>> cora_wd.mat also loads on 3.6.1, I downloaded the zip from your previous
>> given link
>> http://code.google.com/p/tmbp-topicmodel-beliefpropagation/downloads/detail?name=TMBPv1.0.zip&can=2&q=
>>
>> Maybe you should simply update your octave (I tested on GNU/Linux
>> openSUSE 12.1 64bit)
>>
>> octave:1> load cora_wd.mat
>> octave:2> whos
>> Variables in the current scope:
>>
>>   Attr Name         Size                     Bytes  Class
>>   ==== ====         ====                     =====  =====
>>        cora_wd   2961x2410                 1254032  double
>>
>> Total is 103699 elements using 1254032 bytes
>>
>> octave:3> typeinfo(cora_wd)
>> ans = sparse matrix
>>
>>