Question

我想创建具有层次关系的复杂数据图。为简单起见，我创建了有关问题的小样本数据。

样本数据

DS      PNL        Type           Load
DS-1    PNL-101    Normal Power   Room 1, Room 2, Room 3
DS-1    PNL-102    Normal Power   Room 4, Room 5
DS-1    PNL-103    Emergency      Room 6, Room 7
DS-1    PNL-104    Emergency      Room 8
DS-2    PNL-201    Emergency      Office 1, Office 2
DS-2    PNL-202    Normal Power   Office 3, Workshop
DS-2    PNL-203    Light          Corridors
DS-3    PNL-301    Normal Power   Equipment 1, Equipment 2
DS-4    None       Normal Power   Road

数据在excel中，我已将其作为pandas数据框导入。我想创建一个图形，使它显示连接到“ DS”的“ PNL”和连接到每个“ PNL”的“负载”。我需要用定界符“，”分隔“负载”列，以获取单独的负载，例如DS-1至PNL-101的第1室，第2室，第3室。

我尝试了以下代码，在某种程度上可以完成这项工作。它创建了图，但是问题是每个PNL都连接到每个Load（这是错误的）。

例如，如果我要为包含DS-2的所有实例绘制图形，则预期图形应为DS-2与PNL-202，PNL-202和PNL-203连接（我已经实现了）。同样，PNL-201仅应与Office 1和Office 2连接。PNL-202应与Office 3和Workshop连接。 PNL-203应该仅与走廊连接。

import numpy as np
import pandas as pd
import matplotlib as plt
import networkx as nx
#import pygraphviz as pgv
from graphviz import Graph, Digraph
import graphviz # https://www.graphviz.org/doc/info/shapes.html

df = pd.read_excel('data.xlsx')

g = Graph(format='png')

def show_impact(item):
    '''
    Input: The partial string to search in the dataframe df.
           All the rows which contains the string "item", are extracted to 
           form another dataframe df0.
    Output: Plot the graph of the extracted small dataframe.
    '''
    # Search all rows that contains the string "item".
    s = df.stack() # convert entire data frame into a series of values
    df0 = df.iloc[s[s.str.contains(item,na=False)].index.get_level_values(0)]

    # Load column to list of individual locations
    loc = df0['Load'].str.split(',',expand=True).values.tolist()

    # Draw Nodes First
    for i, row in df0.iterrows():
        for j, col in enumerate(df0):

            # Draw DS Nodes
            g.node(str(row['DS']), style='filled', shape='rect', color='#FF5733') 
            # Draw PNL nodes
            g.node(str(row['PNL']), style='filled', shape='rect', color='#FFC300') 
            # Draw Load nodes
    for k in loc:
            for l in k:
                g.node(str(l), style='filled', shape='rect', color='#C0C0C0') 


    # Edges (DS-PNL)
    for i, row in df0.iterrows():
        g.edge(str(row['DS']), str(row['PNL']), color='red', label='')
    # Edges (PNL-Loads)    
    for i, row in df0.iterrows():
        for k in loc:
            for l in k:
                g.edge(str(row['PNL']), str(l), color='red', label='')

    g.view()

show_impact('DS-2')

在Graphviz中使用for循环连接节点

0 个答案: