可視化決策樹（來自scikit-learn的示例）

Question

我是使用sciki-learn的noob，所以請耐心等待。

我正在經歷的例子： http://scikit-learn.org/stable/modules/tree.html#tree

>>> from sklearn.datasets import load_iris 
>>> from sklearn import tree 
>>> iris = load_iris() 
>>> clf = tree.DecisionTreeClassifier() 
>>> clf = clf.fit(iris.data, iris.target) 
>>> from StringIO import StringIO 
>>> out = StringIO() 
>>> out = tree.export_graphviz(clf, out_file=out) 

顯然，graphiz文件就可以使用了。

但是，如何使用graphiz文件繪製樹？ （這個例子並沒有詳細描述樹如何繪製）。

示例代碼和提示比歡迎！

謝謝！

---

更新

我使用Ubuntu 12.04，Python的2.7.3

Answer 1

你跑哪個操作系統？你有沒有安裝graphviz？

在你的榜樣，StringIO()對象，持有Graphviz的數據，這裏有一個方法來檢查數據：

... 
>>> print out.getvalue() 

digraph Tree { 
0 [label="X[2] <= 2.4500\nerror = 0.666667\nsamples = 150\nvalue = [ 50. 50. 50.]", shape="box"] ; 
1 [label="error = 0.0000\nsamples = 50\nvalue = [ 50. 0. 0.]", shape="box"] ; 
0 -> 1 ; 
2 [label="X[3] <= 1.7500\nerror = 0.5\nsamples = 100\nvalue = [ 0. 50. 50.]", shape="box"] ; 
0 -> 2 ; 
3 [label="X[2] <= 4.9500\nerror = 0.168038\nsamples = 54\nvalue = [ 0. 49. 5.]", shape="box"] ; 
2 -> 3 ; 
4 [label="X[3] <= 1.6500\nerror = 0.0407986\nsamples = 48\nvalue = [ 0. 47. 1.]", shape="box"] ; 
3 -> 4 ; 
5 [label="error = 0.0000\nsamples = 47\nvalue = [ 0. 47. 0.]", shape="box"] ; 
4 -> 5 ; 
6 [label="error = 0.0000\nsamples = 1\nvalue = [ 0. 0. 1.]", shape="box"] ; 
4 -> 6 ; 
7 [label="X[3] <= 1.5500\nerror = 0.444444\nsamples = 6\nvalue = [ 0. 2. 4.]", shape="box"] ; 
3 -> 7 ; 
8 [label="error = 0.0000\nsamples = 3\nvalue = [ 0. 0. 3.]", shape="box"] ; 
7 -> 8 ; 
9 [label="X[0] <= 6.9500\nerror = 0.444444\nsamples = 3\nvalue = [ 0. 2. 1.]", shape="box"] ; 
7 -> 9 ; 
10 [label="error = 0.0000\nsamples = 2\nvalue = [ 0. 2. 0.]", shape="box"] ; 
9 -> 10 ; 
11 [label="error = 0.0000\nsamples = 1\nvalue = [ 0. 0. 1.]", shape="box"] ; 
9 -> 11 ; 
12 [label="X[2] <= 4.8500\nerror = 0.0425331\nsamples = 46\nvalue = [ 0. 1. 45.]", shape="box"] ; 
2 -> 12 ; 
13 [label="X[0] <= 5.9500\nerror = 0.444444\nsamples = 3\nvalue = [ 0. 1. 2.]", shape="box"] ; 
12 -> 13 ; 
14 [label="error = 0.0000\nsamples = 1\nvalue = [ 0. 1. 0.]", shape="box"] ; 
13 -> 14 ; 
15 [label="error = 0.0000\nsamples = 2\nvalue = [ 0. 0. 2.]", shape="box"] ; 
13 -> 15 ; 
16 [label="error = 0.0000\nsamples = 43\nvalue = [ 0. 0. 43.]", shape="box"] ; 
12 -> 16 ; 
} 

，你可以把它寫成.dot file併產生圖像的輸出，如源表明您鏈接：

$ dot -Tpng tree.dot -o tree.png（PNG格式輸出）

Answer 2

你非常接近！只是做：

graph_from_dot_data(out.getvalue()).write_pdf("somefile.pdf")