1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
|
%%
%% This is file `nndraw.sty',
%% generated with the docstrip utility.
%%
%% The original source files were:
%%
%% nndraw.dtx (with options: `package')
%%
%% This is a generated file.
%%
%% Copyright (C) 2021 by Carlos Cardoso Dias <carlosdias.dev@gmail.com>
%% --------------------------------------------------------------------------
%% This work may be distributed and/or modified under the
%% conditions of the LaTeX Project Public License, either version 1.3
%% of this license or (at your option) any later version.
%% The latest version of this license is in
%% http://www.latex-project.org/lppl.txt
%% and version 1.3 or later is part of all distributions of LaTeX
%% version 2005/12/01 or later.
%%
\NeedsTeXFormat{LaTeX2e}[1999/12/01]
\ProvidesPackage{nndraw}
[2021/06/22 1.0 nndraw is a LaTeX package which provides utilities to draw neural networks.]
\textbf{nndraw} is a \LaTeX package which provides utilities to draw neural networks in an efficient way.
Currently this package provides utilities to draw fully connected feedforward neural networks with an arbitrary
number of layers described inside the `fullyconnectednn` environment using the command `nnlayer`.
An example of usage is shown below, in this example, a fully connected feedforward
deep neural network is provided with two inputs in the first layer, one hidden layer
with four neurons and one output layer with one output. This example shows how easy it is
to customize the presence/absence of biases in any layer as well as its position.
\begin{verbatim}
\begin{fullyconnectednn}[biasy = -4.5,
titley = 1,
inout = false]
\nnlayer[title = Input layer]{2}
\nnlayer[title = Hidden layer]{4}
\nnlayer[title = Ouput layer, hasbias = false]{1}
\end{fullyconnectednn}
\end{verbatim}
\noindent Which is drawn as
\begin{fullyconnectednn}[biasy = -4.5,
titley = 1,
inout = false]
\nnlayer[title = Input layer]{2}
\nnlayer[title = Hidden layer]{4}
\nnlayer[title = Ouput layer, hasbias = false]{1}
\end{fullyconnectednn}
\NeedsTeXFormat{LaTeX2e}
\ProvidesPackage{nndraw}[2021/06/20 Neural Networks Draw LaTeX package]
\RequirePackage{tikz}
\pgfdeclarelayer{back}
\pgfsetlayers{back,main}
\newcommand\nnlayerNoText[1]{}
\newif\ifnnlayerHasBias
\pgfkeys{
/nnlayer/.is family, /nnlayer,
default/.style = {title = , text = \nnlayerNoText, hasbias = true, bias = , color = , biascolor = },
title/.estore in = \nnlayerTitle,
text/.store in = \nnlayerText,
hasbias/.is if = nnlayerHasBias,
bias/.estore in = \nnlayerBias,
color/.estore in = \nnlayerColor,
biascolor/.estore in = \nnlayerBiasColor,
}
\newcounter{numlayers}
\newcounter{nninputlayer}
\newcounter{lastnnsize}
\tikzstyle{neuron}=[circle, minimum size=6mm, fill=gray!70]
\tikzstyle{neuron_connection}=[->, shorten >=1pt, gray!70]
\newcommand\nnlayer[2][]{
\pgfkeys{/nnlayer, default, #1}
\node[text width = \fullyconnectednnTextWidth, align = center] at (\fullyconnectednnLayersep *
\thenumlayers,\fullyconnectednnTitleY) {\nnlayerTitle};
\if\thenninputlayer0
\setcounter{nninputlayer}{#2}
\fi
\foreach \i in {1,...,#2}
{
\node[neuron, yshift=(#2 - \thenninputlayer)*5 mm, fill=\nnlayerColor] (\thenumlayers-\i) at (\thenumlayers * \fullyconnectednnLayersep,-\i) {\nnlayerText{\i}};
}
\if\thelastnnsize0
\else
\ifnnlayerHasBias
\node[neuron, yshift=5mm, fill=\nnlayerBiasColor] (bias-\thenumlayers) at (\thenumlayers * \fullyconnectednnLayersep -\fullyconnectednnLayersep + \fullyconnectednnBiasX, \fullyconnectednnBiasY) {\nnlayerBias};
\fi
\begin{pgfonlayer}{back}
\foreach \i in {1,...,\thelastnnsize}
{
\foreach \j in {1,...,#2}
{
\draw[neuron_connection] (\the\numexpr\thenumlayers-1\relax-\i) -- (\thenumlayers-\j);
}
}
\ifnnlayerHasBias
\foreach \j in {1,...,#2}
{
\draw[neuron_connection] (bias-\thenumlayers) -- (\thenumlayers-\j);
}
\fi
\end{pgfonlayer}
\fi
\setcounter{lastnnsize}{#2}
\stepcounter{numlayers}
}
\newif\iffullyconnectednnInout
\pgfkeys{
/fullyconnectednn/.is family, /fullyconnectednn,
default/.style = {input = \nnlayerNoText, output = \nnlayerNoText, layersep = 3, biasx = 0.75, biasy = -8.75, titley = 0.75, inout = true, text width = 2cm},
input/.store in = \fullyconnectednnInput,
output/.store in = \fullyconnectednnOutput,
layersep/.estore in = \fullyconnectednnLayersep,
biasx/.estore in = \fullyconnectednnBiasX,
biasy/.estore in = \fullyconnectednnBiasY,
titley/.estore in = \fullyconnectednnTitleY,
inout/.is if = fullyconnectednnInout,
text width/.estore in = \fullyconnectednnTextWidth,
}
\newenvironment{fullyconnectednn}[1][]
{
\pgfkeys{/fullyconnectednn, default, #1}
\begin{center}
\begin{tikzpicture}
\setcounter{numlayers}{0}
\setcounter{nninputlayer}{0}
\setcounter{lastnnsize}{0}
}
{
\iffullyconnectednnInout
\begin{pgfonlayer}{back}
\foreach \i in {1,...,\thenninputlayer}
{
\draw[<-, shorten <=1pt] (0-\i) -- ++(-1,0) node[left]{\fullyconnectednnInput{\i}};
}
\foreach \i in {1,...,\thelastnnsize}
{
\draw[->, shorten <=1pt] (\the\numexpr\thenumlayers-1\relax-\i) -- ++(1,0) node[right]{\fullyconnectednnOutput{\i}};
}
\end{pgfonlayer}
\fi
\end{tikzpicture}
\end{center}
}
\endinput
%%
%% End of file `nndraw.sty'.
|
