This is a simple encapsulation for networkX, which further simplifies the process of networkX for drawing neural network diagrams, so that the author can only focus on the logical writing of neural networks, without spending time on the visual presentation of neural networks. At the same time, this module also inherits all the methods of network and does not affect the normal function calls of networkX.
pip install NetworkXsimple
graph.addNode(
name="node name",
pos=(int layer, int node No) ,
nexts=[
{
"node": "G",
"label":"edge label"
}],
previous=[
{
"node": "G",
"label": "edge label G"
}
],
label="AAAAA",
label_color="pink")
- nexts : output degress dict
- previous : input degress dict
{ "node": "next node name", "label":"edge desc" }
from package_xskj_NetworkXsimple import netGraph
from net import netGraph
# ็คบไพ็จๆณ
graph = netGraph(type=1)
# ๆทปๅ ่็น
graph.addNode(
name="A",
pos=(1, 1),
nexts=[
{
"node": "G",
"label": "edge label"
}],
previous=[
{
"node": "G",
"label": "edge label G"
}
],
label="AAAAA",
label_color="pink")
graph.addNode(name="B", label="BBBBB", pos=(1, 2))
graph.addNode(name="C", label="BBBBB", pos=(1, 3))
graph.addNode(name="D", label="่็นA", pos=(1, 4))
graph.addNode(name="E", label="่็นA", pos=(1, 5))
graph.addNode(name="F", label="่็นA", pos=(1, 6))
graph.addNode(name="I", label="่็นA", pos=(1, 7))
graph.addNode(name="G", label="่็นB", pos=(2, 1))
# graph.addNode(name="H", desc="่็นB", pos=(2, 2))
# graph.addNode(name="Z", desc="่็นB", pos=(2, 3))
graph.addNode(name="1", label="่็นB", pos=(3, 1))
# graph.addNode(name="2", desc="่็นB", pos=(3, 2))
# graph.addNode(name="3", desc="่็นB", pos=(3, 3))
# ๅขๅ ่พน
graph.addEdge(("B", "G"))
# ็ปๅถ็ฝ็ปๅพ
graph.draw()
