無法使用 graphviz 渲染決策樹

Question

我整天都在用這個把頭撞在牆上。 我的導師發布的代碼似乎也不起作用......談論令人沮喪的事情。 這是我在當前 state 中的代碼。 我不知所措。 我能夠構建一個混淆矩陣並獲得 model 的准確性，但嘗試渲染決策樹的運氣為零。 請參閱我的代碼中的最后一段注釋以了解我的問題的具體情況。 在這一點上，任何和所有的幫助都將如此驚人。 先感謝您！

    # -*- coding: utf-8 -*-
"""
Spyder Editor

"""
import pandas
from sklearn.model_selection import train_test_split
from sklearn.tree import DecisionTreeClassifier
import sklearn.metrics
import os
from sklearn import tree
from io import StringIO
from IPython.display import Image
import pydotplus
import graphviz

os.chdir("..")

# Set Display options
pandas.set_option('display.max_columns', None)
pandas.set_option('display.max_rows', None)

# Read in the Mars Crater Data
print("reading data set...")
data = pandas.read_csv('marscrater2_pds.csv', low_memory=(False))
print("cleaning data...")
data_clean = data.dropna()
data_clean['DIAM_CIRCLE_IMAGE'] = pandas.to_numeric(data_clean['DIAM_CIRCLE_IMAGE'], downcast="float")
data_clean['DEPTH_RIMFLOOR_TOPOG'] = pandas.to_numeric(data_clean['DEPTH_RIMFLOOR_TOPOG'], downcast="float")
data_clean['LATITUDE_CIRCLE_IMAGE'] = pandas.to_numeric(data_clean['LATITUDE_CIRCLE_IMAGE'], downcast="float")
data_clean['LONGITUDE_CIRCLE_IMAGE'] = pandas.to_numeric(data_clean['LONGITUDE_CIRCLE_IMAGE'], downcast="float")
data_clean['NUMBER_LAYERS'] = pandas.to_numeric(data_clean['NUMBER_LAYERS'], downcast="integer")

data_clean.dtypes
data_clean.describe()

predictors = data_clean[['LATITUDE_CIRCLE_IMAGE', 'LONGITUDE_CIRCLE_IMAGE',
                         'DIAM_CIRCLE_IMAGE', 'NUMBER_LAYERS']]

print("converting target into a binary categorical variable...")
def OLD(row):
    if row['DEPTH_RIMFLOOR_TOPOG'] > 0.25:
        return 1
    else:
        return 0
    
# apply that changes   
data_clean['IS_DEEP'] = data_clean.apply(lambda row: OLD(row), axis=1)
targets = data_clean["IS_DEEP"]

pred_train, pred_test, tar_train, tar_test = train_test_split(predictors, targets, test_size=0.4)

pred_train.shape
pred_test.shape
tar_train.shape
tar_test.shape

classifier=DecisionTreeClassifier()
classifier=classifier.fit(pred_train, tar_train)
predictions=classifier.predict(pred_test)
print("building model...\n")
print("Confusion Matrix")
print(sklearn.metrics.confusion_matrix(tar_test,predictions))
print("\nAccuracy Score")
print(sklearn.metrics.accuracy_score(tar_test,predictions))

#displaying the tree
# This is where I am stuck.
# Can someone please help me render the decision tree?
# this is the code my instructor used in her example and it doesn't work.
# I've seen many variants but none seem to work.  Please bare in mind I am
# a student.  This is all new to me and I am new to stackoverflow (for those
# of you that take offense to the way I asked my question)
out = StringIO()
tree.export_graphviz(classifier, out_file=out)
graph=pydotplus.graph_from_dot_data(out.getvalue())
Image(graph.create_png())

這是我從控制台看到的 output：

Confusion Matrix
[[132697   5324]
 [  5168  10549]]

Accuracy Score
0.931754023078224
Traceback (most recent call last):

  File "C:\Users\<username>\OneDrive\Documents\IT Nano - Data Analyst\Machine Learning\ML (Week 1).py", line 79, in <module>
    Image(graph.create_png())

  File "C:\ProgramData\Anaconda3\lib\site-packages\pydotplus\graphviz.py", line 1797, in <lambda>
    lambda f=frmt, prog=self.prog: self.create(format=f, prog=prog)

  File "C:\ProgramData\Anaconda3\lib\site-packages\pydotplus\graphviz.py", line 2030, in create
    raise InvocationException(

InvocationException: Program terminated with status: 1. stderr follows: Format: "png" not recognized. Use one of:

Answer 1

我自己解決了這個問題...

#displaying the tree
out = StringIO()
tree.export_graphviz(classifier, out_file='output.dot')
dot_data = export_graphviz(classifier, out_file=None)
graph=pydotplus.graph_from_dot_data(dot_data)
graph.write_png("output.png")