Skip to content

Commit f2a1968

Browse files
a-martyna-martyn
committed
ch8 labs
1 parent 5805734 commit f2a1968

4 files changed

+2260
-3
lines changed

‎Source.gv

+244
Original file line numberDiff line numberDiff line change
@@ -0,0 +1,244 @@
1+
digraph Tree {
2+
node [shape=box] ;
3+
0 [label="X[1] <= 0.5\ngini = 0.484\nsamples = 400\nvalue = [236, 164]"] ;
4+
1 [label="X[9] <= 92.5\ngini = 0.429\nsamples = 315\nvalue = [217, 98]"] ;
5+
0 -> 1 [labeldistance=2.5, labelangle=45, headlabel="True"] ;
6+
2 [label="X[6] <= 57.0\ngini = 0.423\nsamples = 46\nvalue = [14, 32]"] ;
7+
1 -> 2 ;
8+
3 [label="X[2] <= 0.5\ngini = 0.42\nsamples = 10\nvalue = [7, 3]"] ;
9+
2 -> 3 ;
10+
4 [label="gini = 0.0\nsamples = 7\nvalue = [7, 0]"] ;
11+
3 -> 4 ;
12+
5 [label="gini = 0.0\nsamples = 3\nvalue = [0, 3]"] ;
13+
3 -> 5 ;
14+
6 [label="X[8] <= 207.5\ngini = 0.313\nsamples = 36\nvalue = [7, 29]"] ;
15+
2 -> 6 ;
16+
7 [label="X[10] <= 56.5\ngini = 0.469\nsamples = 16\nvalue = [6, 10]"] ;
17+
6 -> 7 ;
18+
8 [label="X[9] <= 88.0\ngini = 0.219\nsamples = 8\nvalue = [1, 7]"] ;
19+
7 -> 8 ;
20+
9 [label="gini = 0.0\nsamples = 7\nvalue = [0, 7]"] ;
21+
8 -> 9 ;
22+
10 [label="gini = 0.0\nsamples = 1\nvalue = [1, 0]"] ;
23+
8 -> 10 ;
24+
11 [label="X[8] <= 38.5\ngini = 0.469\nsamples = 8\nvalue = [5, 3]"] ;
25+
7 -> 11 ;
26+
12 [label="gini = 0.0\nsamples = 2\nvalue = [0, 2]"] ;
27+
11 -> 12 ;
28+
13 [label="X[9] <= 68.5\ngini = 0.278\nsamples = 6\nvalue = [5, 1]"] ;
29+
11 -> 13 ;
30+
14 [label="gini = 0.0\nsamples = 1\nvalue = [0, 1]"] ;
31+
13 -> 14 ;
32+
15 [label="gini = 0.0\nsamples = 5\nvalue = [5, 0]"] ;
33+
13 -> 15 ;
34+
16 [label="X[6] <= 115.0\ngini = 0.095\nsamples = 20\nvalue = [1, 19]"] ;
35+
6 -> 16 ;
36+
17 [label="gini = 0.0\nsamples = 19\nvalue = [0, 19]"] ;
37+
16 -> 17 ;
38+
18 [label="gini = 0.0\nsamples = 1\nvalue = [1, 0]"] ;
39+
16 -> 18 ;
40+
19 [label="X[7] <= 13.5\ngini = 0.37\nsamples = 269\nvalue = [203, 66]"] ;
41+
1 -> 19 ;
42+
20 [label="X[5] <= 124.5\ngini = 0.299\nsamples = 224\nvalue = [183, 41]"] ;
43+
19 -> 20 ;
44+
21 [label="X[9] <= 106.5\ngini = 0.117\nsamples = 96\nvalue = [90, 6]"] ;
45+
20 -> 21 ;
46+
22 [label="X[8] <= 177.0\ngini = 0.266\nsamples = 38\nvalue = [32, 6]"] ;
47+
21 -> 22 ;
48+
23 [label="X[7] <= 0.5\ngini = 0.486\nsamples = 12\nvalue = [7, 5]"] ;
49+
22 -> 23 ;
50+
24 [label="gini = 0.0\nsamples = 6\nvalue = [6, 0]"] ;
51+
23 -> 24 ;
52+
25 [label="X[10] <= 70.5\ngini = 0.278\nsamples = 6\nvalue = [1, 5]"] ;
53+
23 -> 25 ;
54+
26 [label="gini = 0.0\nsamples = 5\nvalue = [0, 5]"] ;
55+
25 -> 26 ;
56+
27 [label="gini = 0.0\nsamples = 1\nvalue = [1, 0]"] ;
57+
25 -> 27 ;
58+
28 [label="X[11] <= 10.5\ngini = 0.074\nsamples = 26\nvalue = [25, 1]"] ;
59+
22 -> 28 ;
60+
29 [label="X[10] <= 66.5\ngini = 0.444\nsamples = 3\nvalue = [2, 1]"] ;
61+
28 -> 29 ;
62+
30 [label="gini = 0.0\nsamples = 1\nvalue = [0, 1]"] ;
63+
29 -> 30 ;
64+
31 [label="gini = 0.0\nsamples = 2\nvalue = [2, 0]"] ;
65+
29 -> 31 ;
66+
32 [label="gini = 0.0\nsamples = 23\nvalue = [23, 0]"] ;
67+
28 -> 32 ;
68+
33 [label="gini = 0.0\nsamples = 58\nvalue = [58, 0]"] ;
69+
21 -> 33 ;
70+
34 [label="X[9] <= 109.5\ngini = 0.397\nsamples = 128\nvalue = [93, 35]"] ;
71+
20 -> 34 ;
72+
35 [label="X[2] <= 0.5\ngini = 0.408\nsamples = 21\nvalue = [6, 15]"] ;
73+
34 -> 35 ;
74+
36 [label="gini = 0.0\nsamples = 5\nvalue = [5, 0]"] ;
75+
35 -> 36 ;
76+
37 [label="X[5] <= 127.5\ngini = 0.117\nsamples = 16\nvalue = [1, 15]"] ;
77+
35 -> 37 ;
78+
38 [label="X[5] <= 126.0\ngini = 0.5\nsamples = 2\nvalue = [1, 1]"] ;
79+
37 -> 38 ;
80+
39 [label="gini = 0.0\nsamples = 1\nvalue = [0, 1]"] ;
81+
38 -> 39 ;
82+
40 [label="gini = 0.0\nsamples = 1\nvalue = [1, 0]"] ;
83+
38 -> 40 ;
84+
41 [label="gini = 0.0\nsamples = 14\nvalue = [0, 14]"] ;
85+
37 -> 41 ;
86+
42 [label="X[9] <= 126.5\ngini = 0.304\nsamples = 107\nvalue = [87, 20]"] ;
87+
34 -> 42 ;
88+
43 [label="X[10] <= 49.5\ngini = 0.444\nsamples = 42\nvalue = [28, 14]"] ;
89+
42 -> 43 ;
90+
44 [label="X[5] <= 137.0\ngini = 0.48\nsamples = 20\nvalue = [8, 12]"] ;
91+
43 -> 44 ;
92+
45 [label="X[6] <= 95.0\ngini = 0.49\nsamples = 14\nvalue = [8, 6]"] ;
93+
44 -> 45 ;
94+
46 [label="X[8] <= 194.0\ngini = 0.397\nsamples = 11\nvalue = [8, 3]"] ;
95+
45 -> 46 ;
96+
47 [label="X[6] <= 72.0\ngini = 0.5\nsamples = 6\nvalue = [3, 3]"] ;
97+
46 -> 47 ;
98+
48 [label="gini = 0.0\nsamples = 3\nvalue = [0, 3]"] ;
99+
47 -> 48 ;
100+
49 [label="gini = 0.0\nsamples = 3\nvalue = [3, 0]"] ;
101+
47 -> 49 ;
102+
50 [label="gini = 0.0\nsamples = 5\nvalue = [5, 0]"] ;
103+
46 -> 50 ;
104+
51 [label="gini = 0.0\nsamples = 3\nvalue = [0, 3]"] ;
105+
45 -> 51 ;
106+
52 [label="gini = 0.0\nsamples = 6\nvalue = [0, 6]"] ;
107+
44 -> 52 ;
108+
53 [label="X[7] <= 11.0\ngini = 0.165\nsamples = 22\nvalue = [20, 2]"] ;
109+
43 -> 53 ;
110+
54 [label="X[9] <= 125.0\ngini = 0.095\nsamples = 20\nvalue = [19, 1]"] ;
111+
53 -> 54 ;
112+
55 [label="gini = 0.0\nsamples = 18\nvalue = [18, 0]"] ;
113+
54 -> 55 ;
114+
56 [label="X[5] <= 142.5\ngini = 0.5\nsamples = 2\nvalue = [1, 1]"] ;
115+
54 -> 56 ;
116+
57 [label="gini = 0.0\nsamples = 1\nvalue = [1, 0]"] ;
117+
56 -> 57 ;
118+
58 [label="gini = 0.0\nsamples = 1\nvalue = [0, 1]"] ;
119+
56 -> 58 ;
120+
59 [label="X[7] <= 12.5\ngini = 0.5\nsamples = 2\nvalue = [1, 1]"] ;
121+
53 -> 59 ;
122+
60 [label="gini = 0.0\nsamples = 1\nvalue = [0, 1]"] ;
123+
59 -> 60 ;
124+
61 [label="gini = 0.0\nsamples = 1\nvalue = [1, 0]"] ;
125+
59 -> 61 ;
126+
62 [label="X[5] <= 147.5\ngini = 0.168\nsamples = 65\nvalue = [59, 6]"] ;
127+
42 -> 62 ;
128+
63 [label="X[8] <= 97.5\ngini = 0.039\nsamples = 50\nvalue = [49, 1]"] ;
129+
62 -> 63 ;
130+
64 [label="X[9] <= 130.5\ngini = 0.219\nsamples = 8\nvalue = [7, 1]"] ;
131+
63 -> 64 ;
132+
65 [label="gini = 0.0\nsamples = 1\nvalue = [0, 1]"] ;
133+
64 -> 65 ;
134+
66 [label="gini = 0.0\nsamples = 7\nvalue = [7, 0]"] ;
135+
64 -> 66 ;
136+
67 [label="gini = 0.0\nsamples = 42\nvalue = [42, 0]"] ;
137+
63 -> 67 ;
138+
68 [label="X[9] <= 147.0\ngini = 0.444\nsamples = 15\nvalue = [10, 5]"] ;
139+
62 -> 68 ;
140+
69 [label="X[2] <= 0.5\ngini = 0.469\nsamples = 8\nvalue = [3, 5]"] ;
141+
68 -> 69 ;
142+
70 [label="gini = 0.0\nsamples = 2\nvalue = [2, 0]"] ;
143+
69 -> 70 ;
144+
71 [label="X[8] <= 172.5\ngini = 0.278\nsamples = 6\nvalue = [1, 5]"] ;
145+
69 -> 71 ;
146+
72 [label="X[9] <= 138.5\ngini = 0.5\nsamples = 2\nvalue = [1, 1]"] ;
147+
71 -> 72 ;
148+
73 [label="gini = 0.0\nsamples = 1\nvalue = [1, 0]"] ;
149+
72 -> 73 ;
150+
74 [label="gini = 0.0\nsamples = 1\nvalue = [0, 1]"] ;
151+
72 -> 74 ;
152+
75 [label="gini = 0.0\nsamples = 4\nvalue = [0, 4]"] ;
153+
71 -> 75 ;
154+
76 [label="gini = 0.0\nsamples = 7\nvalue = [7, 0]"] ;
155+
68 -> 76 ;
156+
77 [label="X[10] <= 54.5\ngini = 0.494\nsamples = 45\nvalue = [20, 25]"] ;
157+
19 -> 77 ;
158+
78 [label="X[5] <= 130.5\ngini = 0.32\nsamples = 25\nvalue = [5, 20]"] ;
159+
77 -> 78 ;
160+
79 [label="X[6] <= 100.0\ngini = 0.459\nsamples = 14\nvalue = [5, 9]"] ;
161+
78 -> 79 ;
162+
80 [label="X[5] <= 113.5\ngini = 0.494\nsamples = 9\nvalue = [5, 4]"] ;
163+
79 -> 80 ;
164+
81 [label="gini = 0.0\nsamples = 2\nvalue = [2, 0]"] ;
165+
80 -> 81 ;
166+
82 [label="X[2] <= 0.5\ngini = 0.49\nsamples = 7\nvalue = [3, 4]"] ;
167+
80 -> 82 ;
168+
83 [label="gini = 0.0\nsamples = 2\nvalue = [2, 0]"] ;
169+
82 -> 83 ;
170+
84 [label="X[11] <= 17.0\ngini = 0.32\nsamples = 5\nvalue = [1, 4]"] ;
171+
82 -> 84 ;
172+
85 [label="gini = 0.0\nsamples = 4\nvalue = [0, 4]"] ;
173+
84 -> 85 ;
174+
86 [label="gini = 0.0\nsamples = 1\nvalue = [1, 0]"] ;
175+
84 -> 86 ;
176+
87 [label="gini = 0.0\nsamples = 5\nvalue = [0, 5]"] ;
177+
79 -> 87 ;
178+
88 [label="gini = 0.0\nsamples = 11\nvalue = [0, 11]"] ;
179+
78 -> 88 ;
180+
89 [label="X[5] <= 122.5\ngini = 0.375\nsamples = 20\nvalue = [15, 5]"] ;
181+
77 -> 89 ;
182+
90 [label="gini = 0.0\nsamples = 10\nvalue = [10, 0]"] ;
183+
89 -> 90 ;
184+
91 [label="X[9] <= 125.0\ngini = 0.5\nsamples = 10\nvalue = [5, 5]"] ;
185+
89 -> 91 ;
186+
92 [label="gini = 0.0\nsamples = 5\nvalue = [0, 5]"] ;
187+
91 -> 92 ;
188+
93 [label="gini = 0.0\nsamples = 5\nvalue = [5, 0]"] ;
189+
91 -> 93 ;
190+
94 [label="X[9] <= 142.5\ngini = 0.347\nsamples = 85\nvalue = [19, 66]"] ;
191+
0 -> 94 [labeldistance=2.5, labelangle=-45, headlabel="False"] ;
192+
95 [label="X[6] <= 34.5\ngini = 0.236\nsamples = 73\nvalue = [10, 63]"] ;
193+
94 -> 95 ;
194+
96 [label="X[9] <= 109.5\ngini = 0.497\nsamples = 13\nvalue = [6, 7]"] ;
195+
95 -> 96 ;
196+
97 [label="gini = 0.0\nsamples = 6\nvalue = [0, 6]"] ;
197+
96 -> 97 ;
198+
98 [label="X[6] <= 22.5\ngini = 0.245\nsamples = 7\nvalue = [6, 1]"] ;
199+
96 -> 98 ;
200+
99 [label="gini = 0.0\nsamples = 1\nvalue = [0, 1]"] ;
201+
98 -> 99 ;
202+
100 [label="gini = 0.0\nsamples = 6\nvalue = [6, 0]"] ;
203+
98 -> 100 ;
204+
101 [label="X[4] <= 0.5\ngini = 0.124\nsamples = 60\nvalue = [4, 56]"] ;
205+
95 -> 101 ;
206+
102 [label="X[9] <= 110.0\ngini = 0.397\nsamples = 11\nvalue = [3, 8]"] ;
207+
101 -> 102 ;
208+
103 [label="gini = 0.0\nsamples = 5\nvalue = [0, 5]"] ;
209+
102 -> 103 ;
210+
104 [label="X[5] <= 129.5\ngini = 0.5\nsamples = 6\nvalue = [3, 3]"] ;
211+
102 -> 104 ;
212+
105 [label="gini = 0.0\nsamples = 3\nvalue = [3, 0]"] ;
213+
104 -> 105 ;
214+
106 [label="gini = 0.0\nsamples = 3\nvalue = [0, 3]"] ;
215+
104 -> 106 ;
216+
107 [label="X[6] <= 98.5\ngini = 0.04\nsamples = 49\nvalue = [1, 48]"] ;
217+
101 -> 107 ;
218+
108 [label="gini = 0.0\nsamples = 40\nvalue = [0, 40]"] ;
219+
107 -> 108 ;
220+
109 [label="X[10] <= 61.5\ngini = 0.198\nsamples = 9\nvalue = [1, 8]"] ;
221+
107 -> 109 ;
222+
110 [label="gini = 0.0\nsamples = 8\nvalue = [0, 8]"] ;
223+
109 -> 110 ;
224+
111 [label="gini = 0.0\nsamples = 1\nvalue = [1, 0]"] ;
225+
109 -> 111 ;
226+
112 [label="X[8] <= 65.5\ngini = 0.375\nsamples = 12\nvalue = [9, 3]"] ;
227+
94 -> 112 ;
228+
113 [label="gini = 0.0\nsamples = 1\nvalue = [0, 1]"] ;
229+
112 -> 113 ;
230+
114 [label="X[8] <= 379.5\ngini = 0.298\nsamples = 11\nvalue = [9, 2]"] ;
231+
112 -> 114 ;
232+
115 [label="X[3] <= 0.5\ngini = 0.18\nsamples = 10\nvalue = [9, 1]"] ;
233+
114 -> 115 ;
234+
116 [label="X[10] <= 34.5\ngini = 0.5\nsamples = 2\nvalue = [1, 1]"] ;
235+
115 -> 116 ;
236+
117 [label="gini = 0.0\nsamples = 1\nvalue = [0, 1]"] ;
237+
116 -> 117 ;
238+
118 [label="gini = 0.0\nsamples = 1\nvalue = [1, 0]"] ;
239+
116 -> 118 ;
240+
119 [label="gini = 0.0\nsamples = 8\nvalue = [8, 0]"] ;
241+
115 -> 119 ;
242+
120 [label="gini = 0.0\nsamples = 1\nvalue = [0, 1]"] ;
243+
114 -> 120 ;
244+
}

‎Source.gv.pdf

23.3 KB
Binary file not shown.

‎ch7_moving_beyond_linearity_applied.ipynb

+3-3
Original file line numberDiff line numberDiff line change
@@ -13,20 +13,20 @@
1313
"import matplotlib.pyplot as plt\n",
1414
"import seaborn as sns\n",
1515
"sns.set()\n",
16+
"\n",
1617
"import statsmodels.formula.api as smf\n",
1718
"import statsmodels.api as sm\n",
1819
"from sklearn.preprocessing import PolynomialFeatures\n",
1920
"from sklearn import linear_model\n",
2021
"from sklearn import metrics\n",
2122
"from sklearn.feature_selection import RFECV\n",
2223
"from sklearn.linear_model import LinearRegression\n",
23-
"import warnings\n",
24-
"from IPython.display import HTML\n",
2524
"\n",
25+
"from IPython.display import HTML\n",
2626
"from ipywidgets import interact\n",
2727
"import ipywidgets as widgets\n",
2828
"import copy\n",
29-
"\n",
29+
"import warnings\n",
3030
"warnings.filterwarnings('ignore')"
3131
]
3232
},

0 commit comments

Comments
 (0)