Das deutsche Python-Forum

Code: Alles auswählen

#!/usr/bin/env python
# -*- utf-8 -*-

import networkx as nx
import numpy as np
import matplotlib.pyplot as plt
import pandas as pd
import os

os.system('clear')
columns=0;
df = pd.read_excel("docs1.xlsx", sheet_name=0)
for zaehler in df: columns=columns+1
rows = df.size/columns
G = nx.DiGraph()

myLists = [[] for i in range(columns)]

for z1 in range(columns):
    for z2 in range(rows):
        ausgabe = df.loc[z2].values[z1]
        myLists[(z1)].append(ausgabe)


'''
# ----------------------------------
# --- Nur einen  Knoten einlesen ---
# ----------------------------------

suchstring='Doc5'
G.add_node(suchstring)
for z in range(columns):
    for j in [j for j,y in enumerate(myLists[(z)]) if y == suchstring]:
        if z != 0:
            #print("j="+str(j)+", zugehoeriges Dokument: "+str(myLists[(0)][j]))
            G.add_edge(myLists[(0)][j], suchstring, weight=0)
'''


# ----------------------------------
# ---    Alle Knoten einlesen    ---
# ----------------------------------

for suchstring in myLists[(0)]:
    #print("Suchstrings: "+str(suchstring))
    G.add_node(suchstring)
    for z in range(columns):
        for j in [j for j,y in enumerate(myLists[(z)]) if y == suchstring]:
            if z != 0:
                #print("j="+str(j)+", zugehoeriges Dokument: "+str(myLists[(0)][j]))
                G.add_edge(myLists[(0)][j], suchstring, weight=0)





# -----------------------------------------------------------------------------------------------

#Graph in entsprechender Form zeichen 
plt.figure(1, figsize=(9,9))
plt.title("====================\nBeziehungen zwischen den Docs\n====================\n")
#nx.draw_graphviz(gr, prog='fdp', node_size = sizes)
#nx.draw_networkx(G, with_labels=True)
#nx.draw_shell(G, with_labels=True, node_size=groesse)
nx.draw_shell(G, with_labels=True, node_size=2000, node_color='#9999FF', edge_color='#FF0000', arrowsize=30)
#nx.draw_spring(G, with_labels=True)
#nx.draw_random(G, with_labels=True)
#nx.draw_circular(G, with_labels=True)
#nx.draw_spectral(G, with_labels=True)
#plt.tick_params(axis='x', which='both', bottom=False, top=False, labelbottom=False)
#plt.tick_params(axis='y', which='both', right=False, left=False, labelleft=False)
for posi in ['right','top','bottom','left']:
    plt.gca().spines[posi].set_visible(False)

#plt.figure(1,figsize=(2000,2000))
#plt.savefig("graph.png", dpi = 1000)

plt.show()

Code: Alles auswählen

#!/usr/bin/env python3
from contextlib import closing
import io
from pathlib import Path

import matplotlib
from matplotlib import pyplot
import networkx
import xlrd
from bottle import (
    abort,
    default_app,
    Jinja2Template,
    jinja2_view as view,
    response,
    route,
    run,
)

matplotlib.use("agg")

SELF_PATH = Path(__file__).absolute().parent
EXCEL_FILE_PATH = SELF_PATH / "docs1.xlsx"

Jinja2Template.defaults = {
    "app": {"title": "Document Graph"},
    "get_url": default_app().get_url,
}


def read_excel(filename):
    with xlrd.open_workbook(filename, ragged_rows=True) as workbook:
        rows = workbook.sheet_by_index(0).get_rows()

        headers = next(rows, None)
        if headers is None:
            raise ValueError("empty document")
        if headers[0].value != "Hauptdokument":
            raise ValueError(
                f"unexpected value in top left cell: {headers[0].value}"
            )

        for row in rows:
            if row and row[0].value != "":
                yield (
                    row[0].value,
                    [cell.value for cell in row[1:] if cell.value],
                )


def check_document_name(name, names):
    if name is not None and name not in names:
        abort(404, f"Document {name!r} not found.")


@route("/image", name="image_all")
@route("/image/<document_name>", name="image")
def get_image(document_name=None):
    with closing(read_excel(EXCEL_FILE_PATH)) as document_references:
        graph = networkx.DiGraph(dict(document_references))

    check_document_name(document_name, graph)

    if document_name is not None:
        graph = networkx.ego_graph(graph, document_name, undirected=True)

    figure = pyplot.figure(1, figsize=(9, 9))
    try:
        networkx.draw_shell(
            graph,
            ax=figure.gca(),
            with_labels=True,
            arrowsize=30,
            node_size=2000,
            node_color="#9999FF",
            edge_color="#FF0000",
        )
        image_file = io.BytesIO()
        figure.savefig(image_file, format="png", dpi=72)
        image_file.seek(0)
        response.set_header("Content-Type", "image/png")
        return image_file
    finally:
        pyplot.close(figure)


@route("/")
@route("/document", name="document_all")
@route("/document/<document_name>", name="document")
@view("index.html")
def index(document_name=None):
    with closing(read_excel(EXCEL_FILE_PATH)) as document_references:
        document_names = [dr[0] for dr in document_references]

    check_document_name(document_name, document_names)

    return {
        "current_document_name": document_name,
        "document_names": document_names,
    }


def main():
    run(reloader=True, debug=True)


if __name__ == "__main__":
    main()

Code: Alles auswählen

<!DOCTYPE html>
<html lang="en">
<head>
  <meta charset="UTF-8">
  <title>{{app.title}}</title>
</head>
<body>
  <h1>
    {{app.title}}
    {%- if current_document_name %} for {{current_document_name}}{% endif %}
  </h1>
  <p>Document names: <strong><a href="{{get_url('document_all')}}">All</a></strong>,
  {% for document_name in document_names -%}
    <a href="{{get_url('document', document_name=document_name)}}">{{document_name}}</a>{{"." if loop.last else ", "}}
  {%- endfor %}</p>
  <img src="{{get_url('image', document_name=current_document_name) if current_document_name else get_url('image_all')}}">
</body>
</html>