usando "if-else" em redes neurais

Question

Eu gostaria de desenhar assim:

Atualmente, o que tenho é

Meu código atual é

\begin{figure}
\centering
\label{fig:nn2}
\begin{tikzpicture}[shorten >=1pt,->,draw=black!50, node distance=\layersep]
    \tikzstyle{every pin edge}=[<-,shorten <=1pt]
    \tikzstyle{neuron}=[circle,fill=black!25,minimum size=17pt,inner sep=0pt]
    \tikzstyle{input neuron}=[neuron, fill=green!50];
    \tikzstyle{output neuron}=[neuron, fill=red!50];
    \tikzstyle{hidden neuron}=[neuron, fill=blue!50];
    \tikzstyle{annot} = [text width=4em, text centered]

    % Draw the input layer nodes
    \foreach \name / \y in {1,...,4}
    % This is the same as writing \foreach \name / \y in {1/1,2/2,3/3,4/4}
        %\if \y in {1,2,3,4}        
        \node[input neuron, pin=left:Input \#\y] (I-\name) at (0,-\y) {$x_{\name}$};
        %\else
        %\node[input neuron, pin=left:Input \#\y] (I-\name) at (0,-\y) {$x_{\name}$};
    %\node[input neuron, pin=left:Input \#4] (+1) at (0,-4) {$x_{\name}$};

    % Draw the hidden layer nodes
    \foreach \name / \y in {1,...,4}
        \path[yshift=0.5cm]
            node[hidden neuron] (H-\name) at (\layersep,-\y cm) {$h_{\name}$};

    % Draw the output layer node
    \node[output neuron,pin={[pin edge={->}]right:Output}, right of=H-3] (O) {$y_{0}$};

    % Connect every node in the input layer with every node in the
    % hidden layer.
    \foreach \source in {1,...,4}
        \foreach \dest in {1,...,4}
            \path (I-\source) edge (H-\dest);

    % Connect every node in the hidden layer with the output layer
    \foreach \source in {1,...,4}
        \path (H-\source) edge (O);

    % Annotate the layers
    \node[annot,above of=H-1, node distance=1cm] (hl) {Hidden layer};
    \node[annot,left of=hl] {Input layer};
    \node[annot,right of=hl] {Output layer};
\end{tikzpicture}
\caption{A figure shows the structure of a general neural networks model}
\end{figure}

Eu tento usar "if-else". e depois de revisar, meu código é:

\begin{figure}
\centering
\label{fig:nn2}
\begin{tikzpicture}[shorten >=1pt,->,draw=black!50, node distance=\layersep]
    \tikzstyle{every pin edge}=[<-,shorten <=1pt]
    \tikzstyle{neuron}=[circle,fill=black!25,minimum size=17pt,inner sep=0pt]
    \tikzstyle{input neuron}=[neuron, fill=green!50];
    \tikzstyle{output neuron}=[neuron, fill=red!50];
    \tikzstyle{hidden neuron}=[neuron, fill=blue!50];
    \tikzstyle{annot} = [text width=4em, text centered]

    % Draw the input layer nodes
    \foreach \name / \y in {1,...,4}
    % This is the same as writing \foreach \name / \y in {1/1,2/2,3/3,4/4}
        \if \y in {1,2,3,4}        
        \node[input neuron, pin=left:Input \#\y] (I-\name) at (0,-\y) {$x_{\name}$};
        \else
        %\node[input neuron, pin=left:Input \#\y] (I-\name) at (0,-\y) {$x_{\name}$};
    %\node[input neuron, pin=left:Input \#4] (+1) at (0,-4) {$x_{\name}$};

    % Draw the hidden layer nodes
    \foreach \name / \y in {1,...,4}
        \path[yshift=0.5cm]
            node[hidden neuron] (H-\name) at (\layersep,-\y cm) {$h_{\name}$};

    % Draw the output layer node
    \node[output neuron,pin={[pin edge={->}]right:Output}, right of=H-3] (O) {$y_{0}$};

    % Connect every node in the input layer with every node in the
    % hidden layer.
    \foreach \source in {1,...,4}
        \foreach \dest in {1,...,4}
            \path (I-\source) edge (H-\dest);

    % Connect every node in the hidden layer with the output layer
    \foreach \source in {1,...,4}
        \path (H-\source) edge (O);

    % Annotate the layers
    \node[annot,above of=H-1, node distance=1cm] (hl) {Hidden layer};
    \node[annot,left of=hl] {Input layer};
    \node[annot,right of=hl] {Output layer};
\end{tikzpicture}
\caption{A figure shows the structure of a general neural networks model}
\end{figure}

no entanto, há um erro: Extra }, ou esquecido \endgroup

Obrigado!

Answer 1

Estavam faltando as chaves ao redor do corpo do loop e \fipara fechar a instrução if. Além disso, a \ifdeclaração não funciona como você pretendia:

\if<token1><token2>(teste se os códigos dos caracteres concordam)

O TeX expandirá as macros a seguir \ifaté que dois tokens não expansíveis sejam encontrados. Se qualquer um dos tokens for uma sequência de controle, o TeX considera que ele possui o código de caractere 256 e o ​​código de categoria 16, a menos que o equivalente atual dessa sequência de controle tenha sido \letigual a um token de caractere não ativo. Desta forma, cada token especifica um par (código de caractere, código de categoria). A condição é verdadeira se os códigos de caracteres forem iguais, independentemente dos códigos de categoria. Por exemplo, depois de \def\a{*}e \let\b=*e \def\c{/}, os testes \if*\ae \if\a\bserão verdadeiros, mas \if\a\cserão falsos. Também \if\a\parserá falso, mas \if\par\letserá verdadeiro.

(O TeXbook página 209)

Portanto você está comparando na primeira etapa da iteração 1e i, na segunda 2e iassim por diante, sempre avaliando como falso. Em vez disso, estou verificando se é a última etapa da iteração com \ifnum:

\begin{figure}
\centering
\label{fig:nn2}
\begin{tikzpicture}[shorten >=1pt,->,draw=black!50, node distance=\layersep]
    \tikzstyle{every pin edge}=[<-,shorten <=1pt]
    \tikzstyle{neuron}=[circle,fill=black!25,minimum size=17pt,inner sep=0pt]
    \tikzstyle{input neuron}=[neuron, fill=green!50];
    \tikzstyle{output neuron}=[neuron, fill=red!50];
    \tikzstyle{hidden neuron}=[neuron, fill=blue!50];
    \tikzstyle{annot} = [text width=4em, text centered]
    \newcommand{\n}{4} % number of neurons per layer

    % Draw the input layer nodes
    \foreach \name / \y in {1,...,\n}{
    % This is the same as writing \foreach \name / \y in {1/1,2/2,3/3,4/4}
        \ifnum \y=\n
            \node[input neuron, pin=left:Input \#$n$] (I-\name) at (0,-\y) {$x_{n}$};
        \else
            \node[input neuron, pin=left:Input \#\y] (I-\name) at (0,-\y) {$x_{\name}$};
        \fi
    }

    % Draw the hidden layer nodes
    \foreach \name / \y in {1,...,\n}{
        \ifnum \y=\n
            \path[yshift=0.5cm] node[hidden neuron] (H-\name) at (\layersep,-\y cm) {$h_{n}$};
        \else
            \path[yshift=0.5cm] node[hidden neuron] (H-\name) at (\layersep,-\y cm) {$h_{\name}$};
        \fi
    }

    % Draw the output layer node
    \node[output neuron,pin={[pin edge={->}]right:Output}, right of=H-3] (O) {$y_{0}$};

    % Connect every node in the input layer with every node in the
    % hidden layer.
    \foreach \source in {1,...,\n}
        \foreach \dest in {1,...,\n}
            \path (I-\source) edge (H-\dest);

    % Connect every node in the hidden layer with the output layer
    \foreach \source in {1,...,\n}
        \path (H-\source) edge (O);

    % Annotate the layers
    \node[annot,above of=H-1, node distance=1cm] (hl) {Hidden layer};
    \node[annot,left of=hl] {Input layer};
    \node[annot,right of=hl] {Output layer};
\end{tikzpicture}
\caption{A figure shows the structure of a general neural networks model}
\end{figure}

No entanto, eu não usaria nenhum if aqui:

\begin{figure}
\centering
\label{fig:nn2}
\begin{tikzpicture}[shorten >=1pt,->,draw=black!50, node distance=\layersep]
    \tikzstyle{every pin edge}=[<-,shorten <=1pt]
    \tikzstyle{neuron}=[circle,fill=black!25,minimum size=17pt,inner sep=0pt]
    \tikzstyle{input neuron}=[neuron, fill=green!50];
    \tikzstyle{output neuron}=[neuron, fill=red!50];
    \tikzstyle{hidden neuron}=[neuron, fill=blue!50];
    \tikzstyle{annot} = [text width=4em, text centered]
    \newcommand{\numberNeuronsPerLayer}{4}
    \edef\numberNeuronsPerLayerMinusOne{\number\numexpr\numberNeuronsPerLayer-1\relax}

    % Draw the input layer nodes
    \foreach \name / \y in {1,...,\numberNeuronsPerLayerMinusOne}{
    % This is the same as writing \foreach \name / \y in {1/1,2/2,3/3,4/4}
        \node[input neuron, pin=left:Input \#\y] (I-\name) at (0,-\y) {$x_{\name}$};
    }
    \node[input neuron, pin=left:Input \#$n$] (I-\numberNeuronsPerLayer) at (0,-\numberNeuronsPerLayer) {$x_{n}$};

    % Draw the hidden layer nodes
    \begin{scope}[yshift=0.5cm]
        \foreach \name / \y in {1,...,\numberNeuronsPerLayerMinusOne}{
            \path node[hidden neuron] (H-\name) at (\layersep,-\y cm) {$h_{\name}$};
        }
        \path node[hidden neuron] (H-\numberNeuronsPerLayer) at (\layersep,-\numberNeuronsPerLayer cm) {$h_{n}$};
    \end{scope}

    % Draw the output layer node
    \node[output neuron,pin={[pin edge={->}]right:Output}, right of=H-3] (O) {$y_{0}$};

    % Connect every node in the input layer with every node in the
    % hidden layer.
    \foreach \source in {1,...,\numberNeuronsPerLayer}
        \foreach \dest in {1,...,\numberNeuronsPerLayer}
            \path (I-\source) edge (H-\dest);

    % Connect every node in the hidden layer with the output layer
    \foreach \source in {1,...,\numberNeuronsPerLayer}
        \path (H-\source) edge (O);

    % Annotate the layers
    \node[annot,above of=H-1, node distance=1cm] (hl) {Hidden layer};
    \node[annot,left of=hl] {Input layer};
    \node[annot,right of=hl] {Output layer};
\end{tikzpicture}
\caption{A figure shows the structure of a general neural networks model}
\end{figure}

Answer 2

Com base na minha resposta à sua pergunta:

\documentclass[tikz, margin=3mm]{standalone}
\usetikzlibrary{calc, chains, positioning}

\begin{document}
    \begin{tikzpicture}[shorten >=1pt,->, draw=black!50,
        node distance = 6mm and 24mm,
          start chain = going below,
every pin edge/.style = {<-,shorten <=1pt},
        neuron/.style = {circle, fill=#1,
                         minimum size=17pt, inner sep=0pt},
         annot/.style = {text width=4em, align=center}
                        ]
% Draw the input and hidden layer nodes
\ifnum\i<4 
    \node[neuron=green!50, on chain,
          pin=180:Input \#\i    % if you not like to have this inputs, just erase them
          ]              (I-\i)  {$x_{\i}$};
    \node[neuron=blue!50,
      right=of I-\i]     (H-\i)  {};
\else 
    \node[neuron=green!50, on chain,
          pin=180:Input \#\i    % if you not like to have this inputs, just erase them
          ]              (I-\i)  {$+1$}; 
    \node[neuron=blue!50,
          right=of I-\i] (H-\i)  {$+1$}; 
\fi
}
% Draw the output layer node
    \node[neuron=red!50,
          right=of $(H-2)!0.5!(H-3)$]  (O-1)   {};
% Connect input nodes with hidden nodes and
%  hiden nodes with output nodes with the output layer
    \foreach \i in {1,...,4}
        \foreach \j in {1,...,4}
{
    \draw (I-\i) edge (H-\j)
          (H-\j) edge (O-1);
}
    \draw (O-1) -- node[below] {$h_{w,b}(x)$} + (2,0);
% Annotate layers
\node[annot,below=of I-4.center]        {Layer 1};
\node[annot,below=of H-4.center]        {Layer 2};
\node[annot,below=of O-1 |- H-4.center] {Layer 3};
    \end{tikzpicture}
\end{document}

Condicionais: como segue acima MWE

\ifnum\i<4
    action 1
\else
    action 2
\fi;

Termo aditivo: seu problema você pode resolver sem declaração condicional:

\documentclass[tikz, margin=3mm]{standalone}
\usetikzlibrary{calc, chains, positioning}

\begin{document}
    \begin{tikzpicture}[shorten >=1pt,->, draw=black!50,
        node distance = 6mm and 24mm,
          start chain = going below,
every pin edge/.style = {<-,shorten <=1pt},
        neuron/.style = {circle, fill=#1,
                         minimum size=17pt, inner sep=0pt},
         annot/.style = {text width=4em, align=center}
                        ]
% Draw the input and hyden layer nodes
\foreach \i in {1,2,3}
{
    \node[neuron=green!50, on chain,
          pin=180:Input \#\i    % if you not like to have this inputs, just erase them
          ]             (I-\i)  {$x_{\i}$};
    \node[neuron=blue!50,
      right=of I-\i]    (H-\i)     {};
}
    \node[neuron=green!50, 
          pin=180:Input \#\i    % if you not like to have this inputs, just erase them
          below=of I-3
          ]              (I-4)  {$+1$};
    \node[neuron=blue!50,
          below=of H-3] (H-4)  {$+1$};
% Draw the output layer node
    \node[neuron=red!50,
          right=of $(H-2)!0.5!(H-3)$]  (O-1)   {};
% Connect input nodes with hidden nodes and
%  hiden nodes with output nodes with the output layer
    \foreach \i in {1,...,4}
        \foreach \j in {1,...,4}
{
    \draw (I-\i) edge (H-\j)
          (H-\j) edge (O-1);
}
    \draw (O-1) -- node[below] {$h_{w,b}(x)$} + (2,0);
% Annotate layers
\node[annot,below=of I-4.center]        {Layer 1};
\node[annot,below=of H-4.center]        {Layer 2};
\node[annot,below=of O-1 |- H-4.center] {Layer 3};
    \end{tikzpicture}
\end{document}

O resultado é o mesmo de antes.