[R-br] [OFF] Biblioteca C para matrizes.
Benilton Carvalho
beniltoncarvalho em gmail.com
Terça Agosto 14 20:49:57 BRT 2012
pq nao a propria GSL?
2012/8/15 Pedro Rafael <pedro.rafael.marinho em gmail.com>:
> Benilton, você conhece alguma biblioteca boa de C++ para trabalhar com
> cálculo diferencial e integral, alguma semelhante à GSL?
>
> Obrigado
> [ ],
> Pedro Rafael Diniz Marinho.
>
>
> Em 13 de agosto de 2012 16:09, Pedro Rafael
> <pedro.rafael.marinho em gmail.com> escreveu:
>> Eu rodei o programa em um ubuntu, instalei o cmake via apt-get e
>> compilei o fonte do armadiilo. O programa rodou sem nenhum problema.
>> Não sei porque no Fedora 17 ele não configurou o arquivo
>> automaticamente. Eu compilei o programa no ubuntu usando: g++ -o
>> programa programa.cpp -larmadillo.
>>
>> Obrigado aí Benilton pela ajuda,
>> Pedro Rafael Diniz Marinho.
>>
>>
>> Em 13 de agosto de 2012 15:32, beniltoncarvalho [via R-br]
>> <ml-node+s2285057n4655857h44 em n4.nabble.com> escreveu:
>>> eh algum config...
>>>
>>> Sent from a mobile device
>>>
>>> On Aug 13, 2012 7:17 PM, "Pedro Rafael" <[hidden email]> wrote:
>>>>
>>>> Vou tentar procurar qual o arquivo que devo fazer essa alteração que
>>>> você me informou.
>>>> Obrigado.
>>>> [ ],
>>>> Pedro Rafael Diniz Marinho.
>>>>
>>>>
>>>> Em 13 de agosto de 2012 05:53, beniltoncarvalho [via R-br]
>>>> <[hidden email]> escreveu:
>>>>
>>>> > mais provavel vc nao ter habilitado LAPACK/BLAS durante a compilacao
>>>> > do Armadillo... ja' nao lembro exatamente o arquivo, mas ha' um q vc
>>>> > precisa modificar manualmente (caso esteja fazendo a instalacao
>>>> > manual) e deixar algo como
>>>> >
>>>> > define ARMA_USE_LAPACK
>>>> > define ARMA_USE_BLAS
>>>> >
>>>> > b
>>>> >
>>>> > 2012/8/13 Pedro Rafael <[hidden email]>:
>>>> >
>>>> >> Será que é alguma coisa com o sistema operacional?
>>>> >> [ ],
>>>> >> Pedro Rafael Diniz Marinho.
>>>> >>
>>>> >>
>>>> >> Em 12 de agosto de 2012 22:36, Pedro Rafael
>>>> >> <[hidden email]> escreveu:
>>>> >>> Tentei remover a biblioteca armadillo e instalar as dependências
>>>> >>> necessárias e no final reinstalei.
>>>> >>> [ ],
>>>> >>> Pedro Rafael Diniz Marinho.
>>>> >>>
>>>> >>>
>>>> >>> Em 12 de agosto de 2012 22:34, Pedro Rafael
>>>> >>> <[hidden email]> escreveu:
>>>> >>>> Valeu Benilton. O problema da compilação n ocorre. Possivelmente as
>>>> >>>> bibliotecas necessárias estão instaladas. Ocorre erro na hora de
>>>> >>>> executar o programa. O erro é:
>>>> >>>>
>>>> >>>> error: det(): use of ATLAS or LAPACK needs to be enabled
>>>> >>>> terminate called after throwing an instance of 'std::logic_error'
>>>> >>>> what():
>>>> >>>> Abortado (imagem do núcleo gravada)
>>>> >>>>
>>>> >>>>
>>>> >>>> [ ],
>>>> >>>> Pedro Rafael Diniz Marinho.
>>>> >>>>
>>>> >>>>
>>>> >>>> Em 12 de agosto de 2012 22:29, beniltoncarvalho [via R-br]
>>>> >>>> <[hidden email]> escreveu:
>>>> >>>>> auto correct fail
>>>> >>>>>
>>>> >>>>> -llapack -lblas
>>>> >>>>>
>>>> >>>>> Sent from a mobile device
>>>> >>>>>
>>>> >>>>> On Aug 13, 2012 2:11 AM, "Pedro Rafael" <[hidden email]> wrote:
>>>> >>>>>>
>>>> >>>>>> Olá Benilton, como vai? O programa que estou testando é bem
>>>> >>>>>> simples.
>>>> >>>>>> Na verdade a matriz é pequena, 5 por 5. Veja:
>>>> >>>>>>
>>>> >>>>>> #include<iostream>
>>>> >>>>>> #include"armadillo"
>>>> >>>>>>
>>>> >>>>>> using namespace arma;
>>>> >>>>>>
>>>> >>>>>> int main()
>>>> >>>>>> {
>>>> >>>>>> mat A = randu<mat>(5,5);
>>>> >>>>>> double x = det(A);
>>>> >>>>>> return 0;
>>>> >>>>>> }
>>>> >>>>>>
>>>> >>>>>> Compilei como você me indicou e obtive o seguinte erro.
>>>> >>>>>>
>>>> >>>>>> /usr/include/armadillo_bits/compiler_setup.hpp:117:6: warning:
>>>> >>>>>> #warning "*** Detected GCC 4.7.0 / 4.7.1, which has a regression
>>>> >>>>>> (bug)" [-Wcpp]
>>>> >>>>>> /usr/include/armadillo_bits/compiler_setup.hpp:118:6: warning:
>>>> >>>>>> #warning "*** See http://gcc.gnu.org/bugzilla/show_bug.cgi?id=53549
>>>> >>>>>> " [-Wcpp]
>>>> >>>>>> /usr/include/armadillo_bits/compiler_setup.hpp:119:6: warning:
>>>> >>>>>> #warning "*** A partial workaround for the bug has been activated,
>>>> >>>>>> " [-Wcpp]
>>>> >>>>>> /usr/include/armadillo_bits/compiler_setup.hpp:120:6: warning:
>>>> >>>>>> #warning "*** which reduces some functionality in fixed-size
>>>> >>>>>> matrices
>>>> >>>>>> " [-Wcpp]
>>>> >>>>>> /usr/bin/ld: cannot find -lapack
>>>> >>>>>> collect2: error: ld returned 1 exit status
>>>> >>>>>>
>>>> >>>>>>
>>>> >>>>>> [ ],
>>>> >>>>>> Pedro Rafael Diniz Marinho.
>>>> >>>>>> Em 12 de agosto de 2012 22:06, beniltoncarvalho [via R-br]
>>>> >>>>>> <[hidden email]> escreveu:
>>>> >>>>>>
>>>> >>>>>> > Vc precisa linkar p lapack blas
>>>> >>>>>> >
>>>> >>>>>> > g++ -O2 -lapack -lblas programa.cpp
>>>> >>>>>> >
>>>> >>>>>> > Obvio, não testado. Adicionalmente, p matriz grande vc deveria
>>>> >>>>>> > usar
>>>> >>>>>> > log_det.
>>>> >>>>>> >
>>>> >>>>>> > Sent from a mobile device
>>>> >>>>>> >
>>>> >>>>>> > On Aug 13, 2012 12:44 AM, "Pedro Rafael" <[hidden email]> wrote:
>>>> >>>>>> >>
>>>> >>>>>> >> Pessoal estou tendo um problema com a biblioteca armadillo no
>>>> >>>>>> >> linux
>>>> >>>>>> >> fedora. Quando vou calcular o determinante de uma matriz muito
>>>> >>>>>> >> pequena
>>>> >>>>>> >> funciona mas quando a minha matriz é grande ocorre o seguinte
>>>> >>>>>> >> erro:
>>>> >>>>>> >>
>>>> >>>>>> >> error: det(): use of ATLAS or LAPACK needs to be enabled
>>>> >>>>>> >>
>>>> >>>>>> >> Já instalei as bibliotecas cmake, blas-devel, lapack-devel,
>>>> >>>>>> >> atlas-devel, boost-devel mas o erro permanece. Alguém já teve
>>>> >>>>>> >> isso
>>>> >>>>>> >> ou
>>>> >>>>>> >> poderia dar uma dica de como corrigir esse problema?
>>>> >>>>>> >>
>>>> >>>>>> >> Obrigado,
>>>> >>>>>> >> Pedro Rafael Diniz Marinho.
>>>> >>>>>> >>
>>>> >>>>>> >>
>>>> >>>>>> >> Em 10 de agosto de 2012 23:58, Pedro Rafael
>>>> >>>>>> >> <[hidden email]> escreveu:
>>>> >>>>>> >>
>>>> >>>>>> >> > Não compreendi os exemplos. Vou dar uma lida maior e buscar
>>>> >>>>>> >> > mais
>>>> >>>>>> >> > coisas
>>>> >>>>>> >> > na net.
>>>> >>>>>> >> >
>>>> >>>>>> >> > Obrigado,
>>>> >>>>>> >> > Pedro Rafael Diniz Marinho.
>>>> >>>>>> >> >
>>>> >>>>>> >> >
>>>> >>>>>> >> > Em 10 de agosto de 2012 23:04, beniltoncarvalho [via R-br]
>>>> >>>>>> >> > <[hidden email]> escreveu:
>>>> >>>>>> >>
>>>> >>>>>> >> >> No básico do basico... Envolva sua função em um extern. Dai
>>>> >>>>>> >> >> use
>>>> >>>>>> >> >> o
>>>> >>>>>> >> >> .Call
>>>> >>>>>> >> >>
>>>> >>>>>> >> >> Sent from a mobile device
>>>> >>>>>> >> >>
>>>> >>>>>> >> >> On Aug 11, 2012 2:42 AM, "Pedro Rafael" <[hidden email]>
>>>> >>>>>> >> >> wrote:
>>>> >>>>>> >> >>>
>>>> >>>>>> >> >>> Interessante. Com o comando RcppArmadillo.package.skeleton é
>>>> >>>>>> >> >>> possível
>>>> >>>>>> >> >>> criar a estrutura de diretórios de um pacote R. Digamos que
>>>> >>>>>> >> >>> eu
>>>> >>>>>> >> >>> tenha
>>>> >>>>>> >> >>> um código C++ e queria arrumar ele para fazer um pacote para
>>>> >>>>>> >> >>> o
>>>> >>>>>> >> >>> R.
>>>> >>>>>> >> >>> Eu
>>>> >>>>>> >> >>> queria entender como se da essa comunicação. Só a filosofia
>>>> >>>>>> >> >>> mesmo.
>>>> >>>>>> >> >>> Eu
>>>> >>>>>> >> >>> entendo como criar um pacote do R utilizando o sistema de
>>>> >>>>>> >> >>> orientação à
>>>> >>>>>> >> >>> objeto S3 e sei que existe outras como S4 e R5. O que não
>>>> >>>>>> >> >>> consigo
>>>> >>>>>> >> >>> entender bem é como comunicar o código C/C++ com o R e como
>>>> >>>>>> >> >>> se
>>>> >>>>>> >> >>> dá a
>>>> >>>>>> >> >>> comunicação desses códigos (C/C++) com o R. Vi que posso
>>>> >>>>>> >> >>> chamar
>>>> >>>>>> >> >>> códigos do R para o C++ mas acho que o mais interessante é o
>>>> >>>>>> >> >>> contrário. Como chamar um código otimizado de C++ para R e
>>>> >>>>>> >> >>> criar
>>>> >>>>>> >> >>> um
>>>> >>>>>> >> >>> pacote caso seja conveniente. Alguém poderia me dar um
>>>> >>>>>> >> >>> exemplo
>>>> >>>>>> >> >>> com
>>>> >>>>>> >> >>> uma
>>>> >>>>>> >> >>> função bem simplória só para entender por alto como
>>>> >>>>>> >> >>> funciona?
>>>> >>>>>> >> >>>
>>>> >>>>>> >> >>> Obrigado,
>>>> >>>>>> >> >>> Pedro Rafael Diniz Marinho.
>>>> >>>>>> >> >>>
>>>> >>>>>> >> >>>
>>>> >>>>>> >> >>> Em 10 de agosto de 2012 22:23, beniltoncarvalho [via R-br]
>>>> >>>>>> >> >>> <[hidden email]> escreveu:
>>>> >>>>>> >> >>>
>>>> >>>>>> >> >>> > Sim. Mas antes de qualquer coisa, veja o pacote
>>>> >>>>>> >> >>> > RcppArmadillo
>>>> >>>>>> >> >>> > e
>>>> >>>>>> >> >>> > os
>>>> >>>>>> >> >>> > documentos de ajuda associados. b
>>>> >>>>>> >> >>> >
>>>> >>>>>> >> >>> > 2012/8/11 Pedro Rafael <[hidden email]>:
>>>> >>>>>> >> >>> >
>>>> >>>>>> >> >>> >> Uma última dúvida. É fácil integrar códigos C++ com o R?
>>>> >>>>>> >> >>> >> Utilizei
>>>> >>>>>> >> >>> >> as
>>>> >>>>>> >> >>> >> funções .C() e .Call() para chamar comandos em C para R.
>>>> >>>>>> >> >>> >> Nunca
>>>> >>>>>> >> >>> >> tentei
>>>> >>>>>> >> >>> >> utilizando C++. A construção de pacotes R feitos
>>>> >>>>>> >> >>> >> utilizando
>>>> >>>>>> >> >>> >> funções
>>>> >>>>>> >> >>> >> escritas em C++ também é tranquila?
>>>> >>>>>> >> >>> >>
>>>> >>>>>> >> >>> >>
>>>> >>>>>> >> >>> >> Obrigado,
>>>> >>>>>> >> >>> >> Pedro Rafael Diniz Marinho.
>>>> >>>>>> >> >>> >>
>>>> >>>>>> >> >>> >>
>>>> >>>>>> >> >>> >> Em 10 de agosto de 2012 22:04, Pedro Rafael
>>>> >>>>>> >> >>> >> <[hidden email]> escreveu:
>>>> >>>>>> >> >>> >>> Que legal Benilton, vou olhar bem esse pacote. Muito
>>>> >>>>>> >> >>> >>> obrigado
>>>> >>>>>> >> >>> >>> pelas
>>>> >>>>>> >> >>> >>> informações. Vou brincar um pouco com esse pacote.
>>>> >>>>>> >> >>> >>> Procurei na
>>>> >>>>>> >> >>> >>> web
>>>> >>>>>> >> >>> >>> outros pacotes de álgebra linear para C++ mas me parece
>>>> >>>>>> >> >>> >>> que
>>>> >>>>>> >> >>> >>> esse é
>>>> >>>>>> >> >>> >>> o
>>>> >>>>>> >> >>> >>> pacote livre mais completo.
>>>> >>>>>> >> >>> >>> [ ],
>>>> >>>>>> >> >>> >>> Pedro Rafael Diniz Marinho.
>>>> >>>>>> >> >>> >>>
>>>> >>>>>> >> >>> >>>
>>>> >>>>>> >> >>> >>> Em 10 de agosto de 2012 21:31, beniltoncarvalho [via
>>>> >>>>>> >> >>> >>> R-br]
>>>> >>>>>> >> >>> >>> <[hidden email]> escreveu:
>>>> >>>>>> >> >>> >>>> Se vc gostou do Armadillo, divirta-se com o pacote
>>>> >>>>>> >> >>> >>>> RcppArmadillo.
>>>> >>>>>> >> >>> >>>> b
>>>> >>>>>> >> >>> >>>>
>>>> >>>>>> >> >>> >>>> 2012/8/11 Pedro Rafael <[hidden email]>:
>>>> >>>>>> >> >>> >>>>
>>>> >>>>>> >> >>> >>>>> Obrigado Prof. Paulo pela informação, me ajudou
>>>> >>>>>> >> >>> >>>>> bastante.
>>>> >>>>>> >> >>> >>>>> Eu
>>>> >>>>>> >> >>> >>>>> gostei
>>>> >>>>>> >> >>> >>>>> muito Benilton da biblioteca de álgebra linear
>>>> >>>>>> >> >>> >>>>> "armadillo".
>>>> >>>>>> >> >>> >>>>> Muito
>>>> >>>>>> >> >>> >>>>> mais
>>>> >>>>>> >> >>> >>>>> fácil que a GSL. Eu acho desorganizado os códigos C. O
>>>> >>>>>> >> >>> >>>>> problema
>>>> >>>>>> >> >>> >>>>> que
>>>> >>>>>> >> >>> >>>>> quando escrevo algum código matemático em C++ sempre
>>>> >>>>>> >> >>> >>>>> tem
>>>> >>>>>> >> >>> >>>>> uma
>>>> >>>>>> >> >>> >>>>> estrutura
>>>> >>>>>> >> >>> >>>>> procedural, deve ser o vício que tenho do C. Acabo
>>>> >>>>>> >> >>> >>>>> usando
>>>> >>>>>> >> >>> >>>>> quase
>>>> >>>>>> >> >>> >>>>> nada
>>>> >>>>>> >> >>> >>>>> de Classes, Heranças, Polimorfismo, etc. Contudo, acho
>>>> >>>>>> >> >>> >>>>> interessante
>>>> >>>>>> >> >>> >>>>> a
>>>> >>>>>> >> >>> >>>>> facilidade de se trabalhar com vetores, matrizes e
>>>> >>>>>> >> >>> >>>>> alocação
>>>> >>>>>> >> >>> >>>>> dinâmica
>>>> >>>>>> >> >>> >>>>> utilizando C++.
>>>> >>>>>> >> >>> >>>>> [ ],
>>>> >>>>>> >> >>> >>>>> Pedro Rafael Diniz Marinho.
>>>> >>>>>> >> >>> >>>>>
>>>> >>>>>> >> >>> >>>>>
>>>> >>>>>> >> >>> >>>>> Em 10 de agosto de 2012 20:28, beniltoncarvalho [via
>>>> >>>>>> >> >>> >>>>> R-br]
>>>> >>>>>> >> >>> >>>>> <[hidden email]> escreveu:
>>>> >>>>>> >> >>> >>>>>> Pedro,
>>>> >>>>>> >> >>> >>>>>>
>>>> >>>>>> >> >>> >>>>>> Há muito deixei de procurar 'a linguagem mais
>>>> >>>>>> >> >>> >>>>>> <escolha o
>>>> >>>>>> >> >>> >>>>>> adjetivo>
>>>> >>>>>> >> >>> >>>>>> para <escolha a tarefa>'. Fico contente em usar
>>>> >>>>>> >> >>> >>>>>> qualquer
>>>> >>>>>> >> >>> >>>>>> linguagem
>>>> >>>>>> >> >>> >>>>>> que
>>>> >>>>>> >> >>> >>>>>> se mostre util e nao me force a reinventar a roda.
>>>> >>>>>> >> >>> >>>>>> Dito
>>>> >>>>>> >> >>> >>>>>> isso,
>>>> >>>>>> >> >>> >>>>>> nao
>>>> >>>>>> >> >>> >>>>>> vejo
>>>> >>>>>> >> >>> >>>>>> la essas diferencas entre C ou C++. E' obvio que
>>>> >>>>>> >> >>> >>>>>> algum
>>>> >>>>>> >> >>> >>>>>> especialista
>>>> >>>>>> >> >>> >>>>>> em
>>>> >>>>>> >> >>> >>>>>> alguma delas encontrara' inumeros casos que servem de
>>>> >>>>>> >> >>> >>>>>> contra-exemplo;
>>>> >>>>>> >> >>> >>>>>> mas enqto ele faz isso, eu sigo adiante resolvendo
>>>> >>>>>> >> >>> >>>>>> outros
>>>> >>>>>> >> >>> >>>>>> problemas.
>>>> >>>>>> >> >>> >>>>>>
>>>> >>>>>> >> >>> >>>>>> b
>>>> >>>>>> >> >>> >>>>>>
>>>> >>>>>> >> >>> >>>>>> 2012/8/10 Paulo Justiniano <[hidden email]>:
>>>> >>>>>> >> >>> >>>>>>
>>>> >>>>>> >> >>> >>>>>>> Pedro
>>>> >>>>>> >> >>> >>>>>>>
>>>> >>>>>> >> >>> >>>>>>> quando se diz em evitar inverter matrizes é porque
>>>> >>>>>> >> >>> >>>>>>> muitas
>>>> >>>>>> >> >>> >>>>>>> contas
>>>> >>>>>> >> >>> >>>>>>> podem
>>>> >>>>>> >> >>> >>>>>>> ser
>>>> >>>>>> >> >>> >>>>>>> feitas sem a inversao explícita
>>>> >>>>>> >> >>> >>>>>>>
>>>> >>>>>> >> >>> >>>>>>> Por exemplo, a conta
>>>> >>>>>> >> >>> >>>>>>>
>>>> >>>>>> >> >>> >>>>>>> X'V^{-1}X
>>>> >>>>>> >> >>> >>>>>>>
>>>> >>>>>> >> >>> >>>>>>> pode ser feita evitando inversao a partir da
>>>> >>>>>> >> >>> >>>>>>> cholesky
>>>> >>>>>> >> >>> >>>>>>> de V
>>>> >>>>>> >> >>> >>>>>>> que pode ser aproveitada de calculos anteriores uma
>>>> >>>>>> >> >>> >>>>>>> vez
>>>> >>>>>> >> >>> >>>>>>> que
>>>> >>>>>> >> >>> >>>>>>> já
>>>> >>>>>> >> >>> >>>>>>> deve
>>>> >>>>>> >> >>> >>>>>>> ter
>>>> >>>>>> >> >>> >>>>>>> sido
>>>> >>>>>> >> >>> >>>>>>> calculada antes para obter as estimativas
>>>> >>>>>> >> >>> >>>>>>>
>>>> >>>>>> >> >>> >>>>>>> No exemplo
>>>> >>>>>> >> >>> >>>>>>>
>>>> >>>>>> >> >>> >>>>>>> cV <- chol(V)
>>>> >>>>>> >> >>> >>>>>>> crossprod(backsolve(cV, X, transpose=T))
>>>> >>>>>> >> >>> >>>>>>>
>>>> >>>>>> >> >>> >>>>>>> isto é ***muito*** mais rapido (compare com
>>>> >>>>>> >> >>> >>>>>>> system.time())
>>>> >>>>>> >> >>> >>>>>>> do que
>>>> >>>>>> >> >>> >>>>>>>
>>>> >>>>>> >> >>> >>>>>>> t(X) %*% solve(V) %*% X
>>>> >>>>>> >> >>> >>>>>>>
>>>> >>>>>> >> >>> >>>>>>>
>>>> >>>>>> >> >>> >>>>>>>
>>>> >>>>>> >> >>> >>>>>>>
>>>> >>>>>> >> >>> >>>>>>> On Thu, 9 Aug 2012, Pedro Rafael wrote:
>>>> >>>>>> >> >>> >>>>>>>
>>>> >>>>>> >> >>> >>>>>>>> Com respeito a evitar de trabalhar com inversas de
>>>> >>>>>> >> >>> >>>>>>>> matrizes é
>>>> >>>>>> >> >>> >>>>>>>> fato
>>>> >>>>>> >> >>> >>>>>>>> mas
>>>> >>>>>> >> >>> >>>>>>>> infelizmente tenho que trabalhar. Para trabalhar
>>>> >>>>>> >> >>> >>>>>>>> com
>>>> >>>>>> >> >>> >>>>>>>> modelos
>>>> >>>>>> >> >>> >>>>>>>> de
>>>> >>>>>> >> >>> >>>>>>>> regressão heterocedásticos vou precisar de cálculo
>>>> >>>>>> >> >>> >>>>>>>> de
>>>> >>>>>> >> >>> >>>>>>>> inversas de
>>>> >>>>>> >> >>> >>>>>>>> matrizes para o cálculo da estrutura de covariância
>>>> >>>>>> >> >>> >>>>>>>> de
>>>> >>>>>> >> >>> >>>>>>>> tal
>>>> >>>>>> >> >>> >>>>>>>> estimador.
>>>> >>>>>> >> >>> >>>>>>>> [ ],
>>>> >>>>>> >> >>> >>>>>>>> Pedro Rafael Diniz Marinho.
>>>> >>>>>> >> >>> >>>>>>>>
>>>> >>>>>> >> >>> >>>>>>>>
>>>> >>>>>> >> >>> >>>>>>>> Em 9 de agosto de 2012 17:02, Pedro Rafael
>>>> >>>>>> >> >>> >>>>>>>> <[hidden email]> escreveu:
>>>> >>>>>> >> >>> >>>>>>>>>
>>>> >>>>>> >> >>> >>>>>>>>> Verdade Benilton inverter matrizes requer um custo
>>>> >>>>>> >> >>> >>>>>>>>> computacional
>>>> >>>>>> >> >>> >>>>>>>>> bastante elevado. Atualmente programo em C e
>>>> >>>>>> >> >>> >>>>>>>>> programei
>>>> >>>>>> >> >>> >>>>>>>>> muito
>>>> >>>>>> >> >>> >>>>>>>>> pouco
>>>> >>>>>> >> >>> >>>>>>>>> em
>>>> >>>>>> >> >>> >>>>>>>>> C++ por curiosidade. Você acreditar que códigos em
>>>> >>>>>> >> >>> >>>>>>>>> C++
>>>> >>>>>> >> >>> >>>>>>>>> perdem
>>>> >>>>>> >> >>> >>>>>>>>> muito
>>>> >>>>>> >> >>> >>>>>>>>> desempenho em comparação à C considerando que a
>>>> >>>>>> >> >>> >>>>>>>>> programação
>>>> >>>>>> >> >>> >>>>>>>>> em
>>>> >>>>>> >> >>> >>>>>>>>> ambas
>>>> >>>>>> >> >>> >>>>>>>>> as linguagens sejam eficientes? Você tem boas
>>>> >>>>>> >> >>> >>>>>>>>> experiências
>>>> >>>>>> >> >>> >>>>>>>>> na
>>>> >>>>>> >> >>> >>>>>>>>> utilização de C++ para programação estatística e
>>>> >>>>>> >> >>> >>>>>>>>> para
>>>> >>>>>> >> >>> >>>>>>>>> uso
>>>> >>>>>> >> >>> >>>>>>>>> em
>>>> >>>>>> >> >>> >>>>>>>>> simulações?
>>>> >>>>>> >> >>> >>>>>>>>>
>>>> >>>>>> >> >>> >>>>>>>>> Gostei bastante dessa biblioteca que você me
>>>> >>>>>> >> >>> >>>>>>>>> passou,
>>>> >>>>>> >> >>> >>>>>>>>> muito
>>>> >>>>>> >> >>> >>>>>>>>> interessante. Não conhecia. A única biblioteca
>>>> >>>>>> >> >>> >>>>>>>>> para
>>>> >>>>>> >> >>> >>>>>>>>> álgebra
>>>> >>>>>> >> >>> >>>>>>>>> linear
>>>> >>>>>> >> >>> >>>>>>>>> livre que eu conhecia era a GSL.
>>>> >>>>>> >> >>> >>>>>>>>>
>>>> >>>>>> >> >>> >>>>>>>>> [ ],
>>>> >>>>>> >> >>> >>>>>>>>> Pedro Rafael Diniz Marinho.
>>>> >>>>>> >> >>> >>>>>>>>>
>>>> >>>>>> >> >>> >>>>>>>>>
>>>> >>>>>> >> >>> >>>>>>>>> Em 9 de agosto de 2012 15:58, beniltoncarvalho
>>>> >>>>>> >> >>> >>>>>>>>> [via
>>>> >>>>>> >> >>> >>>>>>>>> R-br]
>>>> >>>>>> >> >>> >>>>>>>>> <[hidden email]> escreveu:
>>>> >>>>>> >> >>> >>>>>>>>>>
>>>> >>>>>> >> >>> >>>>>>>>>> Eu ainda sou do time "se vc esta' invertendo
>>>> >>>>>> >> >>> >>>>>>>>>> matriz,
>>>> >>>>>> >> >>> >>>>>>>>>> entao
>>>> >>>>>> >> >>> >>>>>>>>>> vc
>>>> >>>>>> >> >>> >>>>>>>>>> esta'
>>>> >>>>>> >> >>> >>>>>>>>>> fazendo algo errado"...
>>>> >>>>>> >> >>> >>>>>>>>>>
>>>> >>>>>> >> >>> >>>>>>>>>> Em C++, vc vai querer olhar o que Armadillo tem a
>>>> >>>>>> >> >>> >>>>>>>>>> oferecer.
>>>> >>>>>> >> >>> >>>>>>>>>>
>>>> >>>>>> >> >>> >>>>>>>>>> b
>>>> >>>>>> >> >>> >>>>>>>>>>
>>>> >>>>>> >> >>> >>>>>>>>>> 2012/8/9 Pedro Rafael <[hidden email]>:
>>>> >>>>>> >> >>> >>>>>>>>>>
>>>> >>>>>> >> >>> >>>>>>>>>>> Pessoal gostaria de saber qual uma boa
>>>> >>>>>> >> >>> >>>>>>>>>>> biblioteca
>>>> >>>>>> >> >>> >>>>>>>>>>> livre
>>>> >>>>>> >> >>> >>>>>>>>>>> de
>>>> >>>>>> >> >>> >>>>>>>>>>> C
>>>> >>>>>> >> >>> >>>>>>>>>>> para
>>>> >>>>>> >> >>> >>>>>>>>>>> trabalhar com operações matriciais como inversas
>>>> >>>>>> >> >>> >>>>>>>>>>> e
>>>> >>>>>> >> >>> >>>>>>>>>>> produtos de
>>>> >>>>>> >> >>> >>>>>>>>>>> matrizes. Qualquer dica será interessante.
>>>> >>>>>> >> >>> >>>>>>>>>>>
>>>> >>>>>> >> >>> >>>>>>>>>>> [ ] Pedro Rafael Diniz Marinho.
>>>> >>>>>> >> >>> >>>>>>>>>>> _______________________________________________
>>>> >>>>>> >> >>> >>>>>>>>>>> R-br mailing list
>>>> >>>>>> >> >>> >>>>>>>>>>> [hidden email]
>>>> >>>>>> >> >>> >>>>>>>>>>>
>>>> >>>>>> >> >>> >>>>>>>>>>>
>>>> >>>>>> >> >>> >>>>>>>>>>>
>>>> >>>>>> >> >>> >>>>>>>>>>> https://listas.inf.ufpr.br/cgi-bin/mailman/listinfo/r-br
>>>> >>>>>> >> >>> >>>>>>>>>>> Leia o guia de postagem
>>>> >>>>>> >> >>> >>>>>>>>>>> (http://www.leg.ufpr.br/r-br-guia)
>>>> >>>>>> >> >>> >>>>>>>>>>> e
>>>> >>>>>> >> >>> >>>>>>>>>>> forneça
>>>> >>>>>> >> >>> >>>>>>>>>>> código mínimo reproduzível.
>>>> >>>>>> >> >>> >>>>>>>>>>
>>>> >>>>>> >> >>> >>>>>>>>>> _______________________________________________
>>>> >>>>>> >> >>> >>>>>>>>>> R-br mailing list
>>>> >>>>>> >> >>> >>>>>>>>>> [hidden email]
>>>> >>>>>> >> >>> >>>>>>>>>>
>>>> >>>>>> >> >>> >>>>>>>>>>
>>>> >>>>>> >> >>> >>>>>>>>>> https://listas.inf.ufpr.br/cgi-bin/mailman/listinfo/r-br
>>>> >>>>>> >> >>> >>>>>>>>>> Leia o guia de postagem
>>>> >>>>>> >> >>> >>>>>>>>>> (http://www.leg.ufpr.br/r-br-guia)
>>>> >>>>>> >> >>> >>>>>>>>>> e
>>>> >>>>>> >> >>> >>>>>>>>>> forneça
>>>> >>>>>> >> >>> >>>>>>>>>> código
>>>> >>>>>> >> >>> >>>>>>>>>> mínimo reproduzível.
>>>> >>>>>> >> >>> >>>>>>>>>>
>>>> >>>>>> >> >>> >>>>>>>>>>
>>>> >>>>>> >> >>> >>>>>>>>>> ________________________________
>>>> >>>>>> >> >>> >>>>>>>>>> If you reply to this email, your message will be
>>>> >>>>>> >> >>> >>>>>>>>>> added
>>>> >>>>>> >> >>> >>>>>>>>>> to
>>>> >>>>>> >> >>> >>>>>>>>>> the
>>>> >>>>>> >> >>> >>>>>>>>>> discussion
>>>> >>>>>> >> >>> >>>>>>>>>> below:
>>>> >>>>>> >> >>> >>>>>>>>>>
>>>> >>>>>> >> >>> >>>>>>>>>>
>>>> >>>>>> >> >>> >>>>>>>>>>
>>>> >>>>>> >> >>> >>>>>>>>>>
>>>> >>>>>> >> >>> >>>>>>>>>>
>>>> >>>>>> >> >>> >>>>>>>>>>
>>>> >>>>>> >> >>> >>>>>>>>>>
>>>> >>>>>> >> >>> >>>>>>>>>>
>>>> >>>>>> >> >>> >>>>>>>>>>
>>>> >>>>>> >> >>> >>>>>>>>>> http://r-br.2285057.n4.nabble.com/R-br-OFF-Biblioteca-C-para-matrizes-tp4655808p4655810.html
>>>> >>>>>> >> >>> >>>>>>>>>> To unsubscribe from R-br, click here.
>>>> >>>>>> >> >>> >>>>>>>>>> NAML
>>>> >>>>>> >> >>> >>>>>>>>
>>>> >>>>>> >> >>> >>>>>>>> _______________________________________________
>>>> >>>>>> >> >>> >>>>>>>> R-br mailing list
>>>> >>>>>> >> >>> >>>>>>>> [hidden email]
>>>> >>>>>> >> >>> >>>>>>>>
>>>> >>>>>> >> >>> >>>>>>>>
>>>> >>>>>> >> >>> >>>>>>>> https://listas.inf.ufpr.br/cgi-bin/mailman/listinfo/r-br
>>>> >>>>>> >> >>> >>>>>>>> Leia o guia de postagem
>>>> >>>>>> >> >>> >>>>>>>> (http://www.leg.ufpr.br/r-br-guia)
>>>> >>>>>> >> >>> >>>>>>>> e
>>>> >>>>>> >> >>> >>>>>>>> forneça
>>>> >>>>>> >> >>> >>>>>>>> código mínimo reproduzível.
>>>> >>>>>> >> >>> >>>>>>>
>>>> >>>>>> >> >>> >>>>>>>
>>>> >>>>>> >> >>> >>>>>>> _______________________________________________
>>>> >>>>>> >> >>> >>>>>>> R-br mailing list
>>>> >>>>>> >> >>> >>>>>>> [hidden email]
>>>> >>>>>> >> >>> >>>>>>>
>>>> >>>>>> >> >>> >>>>>>>
>>>> >>>>>> >> >>> >>>>>>> https://listas.inf.ufpr.br/cgi-bin/mailman/listinfo/r-br
>>>> >>>>>> >> >>> >>>>>>> Leia o guia de postagem
>>>> >>>>>> >> >>> >>>>>>> (http://www.leg.ufpr.br/r-br-guia)
>>>> >>>>>> >> >>> >>>>>>> e
>>>> >>>>>> >> >>> >>>>>>> forneça
>>>> >>>>>> >> >>> >>>>>>> código
>>>> >>>>>> >> >>> >>>>>>> mínimo reproduzível.
>>>> >>>>>> >> >>> >>>>>> _______________________________________________
>>>> >>>>>> >> >>> >>>>>> R-br mailing list
>>>> >>>>>> >> >>> >>>>>> [hidden email]
>>>> >>>>>> >> >>> >>>>>>
>>>> >>>>>> >> >>> >>>>>> https://listas.inf.ufpr.br/cgi-bin/mailman/listinfo/r-br
>>>> >>>>>> >> >>> >>>>>> Leia o guia de postagem
>>>> >>>>>> >> >>> >>>>>> (http://www.leg.ufpr.br/r-br-guia) e
>>>> >>>>>> >> >>> >>>>>> forneça
>>>> >>>>>> >> >>> >>>>>> código
>>>> >>>>>> >> >>> >>>>>> mínimo reproduzível.
>>>> >>>>>> >> >>> >>>>>>
>>>> >>>>>> >> >>> >>>>>>
>>>> >>>>>> >> >>> >>>>>> ________________________________
>>>> >>>>>> >> >>> >>>>>> If you reply to this email, your message will be
>>>> >>>>>> >> >>> >>>>>> added
>>>> >>>>>> >> >>> >>>>>> to
>>>> >>>>>> >> >>> >>>>>> the
>>>> >>>>>> >> >>> >>>>>> discussion
>>>> >>>>>> >> >>> >>>>>> below:
>>>> >>>>>> >> >>> >>>>>>
>>>> >>>>>> >> >>> >>>>>>
>>>> >>>>>> >> >>> >>>>>>
>>>> >>>>>> >> >>> >>>>>>
>>>> >>>>>> >> >>> >>>>>>
>>>> >>>>>> >> >>> >>>>>>
>>>> >>>>>> >> >>> >>>>>>
>>>> >>>>>> >> >>> >>>>>> http://r-br.2285057.n4.nabble.com/R-br-OFF-Biblioteca-C-para-matrizes-tp4655808p4655826.html
>>>> >>>>>> >> >>> >>>>>> To unsubscribe from R-br, click here.
>>>> >>>>>> >> >>> >>>>>> NAML
>>>> >>>>>> >> >>> >>>>> _______________________________________________
>>>> >>>>>> >> >>> >>>>> R-br mailing list
>>>> >>>>>> >> >>> >>>>> [hidden email]
>>>> >>>>>> >> >>> >>>>>
>>>> >>>>>> >> >>> >>>>> https://listas.inf.ufpr.br/cgi-bin/mailman/listinfo/r-br
>>>> >>>>>> >> >>> >>>>> Leia o guia de postagem
>>>> >>>>>> >> >>> >>>>> (http://www.leg.ufpr.br/r-br-guia) e
>>>> >>>>>> >> >>> >>>>> forneça
>>>> >>>>>> >> >>> >>>>> código mínimo reproduzível.
>>>> >>>>>> >> >>> >>>> _______________________________________________
>>>> >>>>>> >> >>> >>>> R-br mailing list
>>>> >>>>>> >> >>> >>>> [hidden email]
>>>> >>>>>> >> >>> >>>>
>>>> >>>>>> >> >>> >>>> https://listas.inf.ufpr.br/cgi-bin/mailman/listinfo/r-br
>>>> >>>>>> >> >>> >>>> Leia o guia de postagem
>>>> >>>>>> >> >>> >>>> (http://www.leg.ufpr.br/r-br-guia)
>>>> >>>>>> >> >>> >>>> e
>>>> >>>>>> >> >>> >>>> forneça
>>>> >>>>>> >> >>> >>>> código
>>>> >>>>>> >> >>> >>>> mínimo reproduzível.
>>>> >>>>>> >> >>> >>>>
>>>> >>>>>> >> >>> >>>>
>>>> >>>>>> >> >>> >>>> ________________________________
>>>> >>>>>> >> >>> >>>> If you reply to this email, your message will be added
>>>> >>>>>> >> >>> >>>> to
>>>> >>>>>> >> >>> >>>> the
>>>> >>>>>> >> >>> >>>> discussion
>>>> >>>>>> >> >>> >>>> below:
>>>> >>>>>> >> >>> >>>>
>>>> >>>>>> >> >>> >>>>
>>>> >>>>>> >> >>> >>>>
>>>> >>>>>> >> >>> >>>>
>>>> >>>>>> >> >>> >>>>
>>>> >>>>>> >> >>> >>>>
>>>> >>>>>> >> >>> >>>> http://r-br.2285057.n4.nabble.com/R-br-OFF-Biblioteca-C-para-matrizes-tp4655808p4655830.html
>>>> >>>>>> >> >>> >>>> To unsubscribe from R-br, click here.
>>>> >>>>>> >> >>> >>>> NAML
>>>> >>>>>> >> >>> >> _______________________________________________
>>>> >>>>>> >> >>> >> R-br mailing list
>>>> >>>>>> >> >>> >> [hidden email]
>>>> >>>>>> >> >>> >> https://listas.inf.ufpr.br/cgi-bin/mailman/listinfo/r-br
>>>> >>>>>> >> >>> >> Leia o guia de postagem
>>>> >>>>>> >> >>> >> (http://www.leg.ufpr.br/r-br-guia) e
>>>> >>>>>> >> >>> >> forneça
>>>> >>>>>> >> >>> >> código mínimo reproduzível.
>>>> >>>>>> >> >>> > _______________________________________________
>>>> >>>>>> >> >>> > R-br mailing list
>>>> >>>>>> >> >>> > [hidden email]
>>>> >>>>>> >> >>> > https://listas.inf.ufpr.br/cgi-bin/mailman/listinfo/r-br
>>>> >>>>>> >> >>> > Leia o guia de postagem (http://www.leg.ufpr.br/r-br-guia)
>>>> >>>>>> >> >>> > e
>>>> >>>>>> >> >>> > forneça
>>>> >>>>>> >> >>> > código
>>>> >>>>>> >> >>> > mínimo reproduzível.
>>>> >>>>>> >> >>> >
>>>> >>>>>> >> >>> >
>>>> >>>>>> >> >>> > ________________________________
>>>> >>>>>> >> >>> > If you reply to this email, your message will be added to
>>>> >>>>>> >> >>> > the
>>>> >>>>>> >> >>> > discussion
>>>> >>>>>> >> >>> > below:
>>>> >>>>>> >> >>> >
>>>> >>>>>> >> >>> >
>>>> >>>>>> >> >>> >
>>>> >>>>>> >> >>> >
>>>> >>>>>> >> >>> >
>>>> >>>>>> >> >>> > http://r-br.2285057.n4.nabble.com/R-br-OFF-Biblioteca-C-para-matrizes-tp4655808p4655833.html
>>>> >>>>>> >> >>> > To unsubscribe from R-br, click here.
>>>> >>>>>> >> >>> > NAML
>>>> >>>>>> >> >>> _______________________________________________
>>>> >>>>>> >> >>> R-br mailing list
>>>> >>>>>> >> >>> [hidden email]
>>>> >>>>>> >> >>> https://listas.inf.ufpr.br/cgi-bin/mailman/listinfo/r-br
>>>> >>>>>> >> >>> Leia o guia de postagem (http://www.leg.ufpr.br/r-br-guia) e
>>>> >>>>>> >> >>> forneça
>>>> >>>>>> >> >>> código mínimo reproduzível.
>>>> >>>>>> >> >>
>>>> >>>>>> >> >>
>>>> >>>>>> >> >> _______________________________________________
>>>> >>>>>> >> >> R-br mailing list
>>>> >>>>>> >> >> [hidden email]
>>>> >>>>>> >> >> https://listas.inf.ufpr.br/cgi-bin/mailman/listinfo/r-br
>>>> >>>>>> >> >> Leia o guia de postagem (http://www.leg.ufpr.br/r-br-guia) e
>>>> >>>>>> >> >> forneça
>>>> >>>>>> >> >> código
>>>> >>>>>> >> >> mínimo reproduzível.
>>>> >>>>>> >> >>
>>>> >>>>>> >> >> ________________________________
>>>> >>>>>> >> >> If you reply to this email, your message will be added to the
>>>> >>>>>> >> >> discussion
>>>> >>>>>> >> >> below:
>>>> >>>>>> >> >>
>>>> >>>>>> >> >>
>>>> >>>>>> >> >>
>>>> >>>>>> >> >>
>>>> >>>>>> >> >> http://r-br.2285057.n4.nabble.com/R-br-OFF-Biblioteca-C-para-matrizes-tp4655808p4655836.html
>>>> >>>>>> >> >> To unsubscribe from R-br, click here.
>>>> >>>>>> >> >> NAML
>>>> >>>>>> >> _______________________________________________
>>>> >>>>>> >> R-br mailing list
>>>> >>>>>> >> [hidden email]
>>>> >>>>>> >> https://listas.inf.ufpr.br/cgi-bin/mailman/listinfo/r-br
>>>> >>>>>> >> Leia o guia de postagem (http://www.leg.ufpr.br/r-br-guia) e
>>>> >>>>>> >> forneça
>>>> >>>>>> >> código mínimo reproduzível.
>>>> >>>>>> >
>>>> >>>>>> >
>>>> >>>>>> > _______________________________________________
>>>> >>>>>> > R-br mailing list
>>>> >>>>>> > [hidden email]
>>>> >>>>>> > https://listas.inf.ufpr.br/cgi-bin/mailman/listinfo/r-br
>>>> >>>>>> > Leia o guia de postagem (http://www.leg.ufpr.br/r-br-guia) e
>>>> >>>>>> > forneça
>>>> >>>>>> > código
>>>> >>>>>> > mínimo reproduzível.
>>>> >>>>>> >
>>>> >>>>>> > ________________________________
>>>> >>>>>> > If you reply to this email, your message will be added to the
>>>> >>>>>> > discussion
>>>> >>>>>> > below:
>>>> >>>>>> >
>>>> >>>>>> >
>>>> >>>>>> >
>>>> >>>>>> > http://r-br.2285057.n4.nabble.com/R-br-OFF-Biblioteca-C-para-matrizes-tp4655808p4655841.html
>>>> >>>>>> > To unsubscribe from R-br, click here.
>>>> >>>>>> > NAML
>>>> >>>>>> _______________________________________________
>>>> >>>>>> R-br mailing list
>>>> >>>>>> [hidden email]
>>>> >>>>>> https://listas.inf.ufpr.br/cgi-bin/mailman/listinfo/r-br
>>>> >>>>>> Leia o guia de postagem (http://www.leg.ufpr.br/r-br-guia) e
>>>> >>>>>> forneça
>>>> >>>>>> código mínimo reproduzível.
>>>> >>>>>
>>>> >>>>>
>>>> >>>>> _______________________________________________
>>>> >>>>> R-br mailing list
>>>> >>>>> [hidden email]
>>>> >>>>> https://listas.inf.ufpr.br/cgi-bin/mailman/listinfo/r-br
>>>> >>>>> Leia o guia de postagem (http://www.leg.ufpr.br/r-br-guia) e forneça
>>>> >>>>> código
>>>> >>>>> mínimo reproduzível.
>>>> >>>>>
>>>> >>>>> ________________________________
>>>> >>>>> If you reply to this email, your message will be added to the
>>>> >>>>> discussion
>>>> >>>>> below:
>>>> >>>>>
>>>> >>>>>
>>>> >>>>> http://r-br.2285057.n4.nabble.com/R-br-OFF-Biblioteca-C-para-matrizes-tp4655808p4655843.html
>>>> >>>>> To unsubscribe from R-br, click here.
>>>> >>>>> NAML
>>>> >> _______________________________________________
>>>> >> R-br mailing list
>>>> >> [hidden email]
>>>> >> https://listas.inf.ufpr.br/cgi-bin/mailman/listinfo/r-br
>>>> >> Leia o guia de postagem (http://www.leg.ufpr.br/r-br-guia) e forneça
>>>> >> código mínimo reproduzível.
>>>> > _______________________________________________
>>>> > R-br mailing list
>>>> > [hidden email]
>>>> > https://listas.inf.ufpr.br/cgi-bin/mailman/listinfo/r-br
>>>> > Leia o guia de postagem (http://www.leg.ufpr.br/r-br-guia) e forneça
>>>> > código
>>>> > mínimo reproduzível.
>>>> >
>>>> >
>>>> > ________________________________
>>>> > If you reply to this email, your message will be added to the discussion
>>>> > below:
>>>> >
>>>> > http://r-br.2285057.n4.nabble.com/R-br-OFF-Biblioteca-C-para-matrizes-tp4655808p4655847.html
>>>> > To unsubscribe from R-br, click here.
>>>> > NAML
>>>> _______________________________________________
>>>> R-br mailing list
>>>> [hidden email]
>>>> https://listas.inf.ufpr.br/cgi-bin/mailman/listinfo/r-br
>>>> Leia o guia de postagem (http://www.leg.ufpr.br/r-br-guia) e forneça
>>>> código mínimo reproduzível.
>>>
>>>
>>> _______________________________________________
>>> R-br mailing list
>>> [hidden email]
>>> https://listas.inf.ufpr.br/cgi-bin/mailman/listinfo/r-br
>>> Leia o guia de postagem (http://www.leg.ufpr.br/r-br-guia) e forneça código
>>> mínimo reproduzível.
>>>
>>> ________________________________
>>> If you reply to this email, your message will be added to the discussion
>>> below:
>>> http://r-br.2285057.n4.nabble.com/R-br-OFF-Biblioteca-C-para-matrizes-tp4655808p4655857.html
>>> To unsubscribe from R-br, click here.
>>> NAML
> _______________________________________________
> R-br mailing list
> R-br em listas.c3sl.ufpr.br
> https://listas.inf.ufpr.br/cgi-bin/mailman/listinfo/r-br
> Leia o guia de postagem (http://www.leg.ufpr.br/r-br-guia) e forneça código mínimo reproduzível.
Mais detalhes sobre a lista de discussão R-br