Quero desenhar um perceptron multicamadas como este:
E criei o seguinte código:
\documentclass[tikz, margin=3mm] {standalone}
\usetikzlibrary{arrows.meta, matrix}
\begin{document}
\begin{tikzpicture}[
> = Stealth, thick,
plain/.style = {draw=none, fill=none, yshift=11mm, text width=7ex, align=center},% for text in images,
ec/.style = {draw=none, fill=none},% for emty cells,
net/.style = {matrix of nodes, % for matrix style
nodes={circle,fill=blue!20, draw, semithick, minimum width=12mm, inner sep=0mm},% circles in image
nodes in empty cells,% for not used cells in matrix
column sep = 16mm, % distance between columns in matrix
row sep = -3mm % distance between rows in matrix
}]
\matrix[net] (m)% m is matrix name, it is used for names of cell: firs has name m-1-1
% in empty space between ampersands will show circles:
% i.e.: nodes of the neural network
{
|[plain]| Input layer & |[plain]| Hidden layer & |[plain]| Output layer \\
|[ec]| & |[ec]| & |[ec]| \\
& -1 & |[ec]| \\
|[ec]| & |[ec]| & |[ec]| \\
|[ec]| & |[ec]| & 3 \\
|[ec]| & |[ec]| & |[ec]| \\
& -1 & |[ec]| \\
|[ec]| & |[ec]| & |[ec]| \\
};
\draw[<-] (m-3-1) -- node[left, xshift=-0.65cm] {$x_1$} +(-2cm,0);
\draw[<-] (m-7-1) -- node[left, xshift=-0.65cm] {$x_2$} +(-2cm,0);
\draw[->] (m-5-3) -- node[right, xshift=0.65cm] {$y$} +(2cm,0);
\draw[->] (m-3-1) -- node[above] {$-2$} (m-3-2);
\draw[->] (m-3-1) -- node[below, yshift=-0.75cm, xshift=-0.5cm] {$2$} (m-7-2);
\draw[->] (m-7-1) -- node[above, yshift=0.75cm, xshift=-0.5cm] {$2$} (m-3-2);
\draw[->] (m-7-1) -- node[below] {$-2$} (m-7-2);
\draw[->] (m-3-2) -- node[above] {$2$} (m-5-3);
\draw[->] (m-7-2) -- node[below] {$2$} (m-5-3);
\end{tikzpicture}
\end{document}
Esta é a saída acima:
O problema é que o desenho do tikz fica assimétrico sempre que removo um nó da primeira camada. Quando desenho um “1” nele, a imagem é simétrica.
Então, como posso consertar isso? Tentei ajustar o tamanho mínimo, mas isso não ajuda.
Obrigado!
Responder1
A matrix of nodesaltera as âncoras do nó interno para base, se você mantê-las center(com nodes={... anchor=center}o diagrama será simétrico novamente.
\documentclass[tikz, margin=3mm] {standalone}
\usetikzlibrary{arrows.meta, matrix}
\begin{document}
\begin{tikzpicture}[
> = Stealth, thick,
plain/.style = {draw=none, fill=none, yshift=11mm, text width=7ex, align=center},% for text in images,
ec/.style = {draw=none, fill=none},% for emty cells,
net/.style = {matrix of nodes, % for matrix style
nodes={circle,fill=blue!20, draw, semithick, minimum width=12mm, inner sep=0mm, anchor=center},% circles in image
nodes in empty cells,% for not used cells in matrix
column sep = 16mm, % distance between columns in matrix
row sep = -3mm % distance between rows in matrix
}]
\matrix[net] (m)% m is matrix name, it is used for names of cell: firs has name m-1-1
% in empty space between ampersands will show circles:
% i.e.: nodes of the neural network
{
|[plain]| Input layer & |[plain]| Hidden layer & |[plain]| Output layer \\
|[ec]| & |[ec]| & |[ec]| \\
& -1 & |[ec]| \\
|[ec]| & |[ec]| & |[ec]| \\
|[ec]| & |[ec]| & 3 \\
|[ec]| & |[ec]| & |[ec]| \\
& -1 & |[ec]| \\
|[ec]| & |[ec]| & |[ec]| \\
};
\draw[<-] (m-3-1) -- node[left, xshift=-0.65cm] {$x_1$} +(-2cm,0);
\draw[<-] (m-7-1) -- node[left, xshift=-0.65cm] {$x_2$} +(-2cm,0);
\draw[->] (m-5-3) -- node[right, xshift=0.65cm] {$y$} +(2cm,0);
\draw[->] (m-3-1) -- node[above] {$-2$} (m-3-2);
\draw[->] (m-3-1) -- node[below, yshift=-0.75cm, xshift=-0.5cm] {$2$} (m-7-2);
\draw[->] (m-7-1) -- node[above, yshift=0.75cm, xshift=-0.5cm] {$2$} (m-3-2);
\draw[->] (m-7-1) -- node[below] {$-2$} (m-7-2);
\draw[->] (m-3-2) -- node[above] {$2$} (m-5-3);
\draw[->] (m-7-2) -- node[below] {$2$} (m-5-3);
\end{tikzpicture}
\end{document}
