7 Decision Trees in Python

Decision Trees¶

In which we talk about decision trees!

In [1]:

import numpy as np
import pandas
import matplotlib.pyplot as plt
%matplotlib inline
import os

from sklearn.model_selection import train_test_split
from sklearn import datasets

##New!
from sklearn import tree



#os.getcwd() ## run if things aren't working as expected!

In [2]:

import sklearn as sklearn

sklearn.__version__

Out[2]:

'1.2.1'

In [3]:

##load our dataset
heloc= pandas.read_csv("data\Heloc_labeled.csv") 

## look it over
heloc.head(10)
#heloc.shape

Out[3]:

.dataframe tbody tr th:only-of-type {
vertical-align: middle;
}

.dataframe tbody tr th {
vertical-align: top;
}

.dataframe thead th {
text-align: right;
}

	Age	Sex	Income	HELOC
0	30	Female	101000	0
1	25	Male	86000	0
2	20	Male	50000	0
3	26	Male	58000	0
4	18	Female	93000	0
5	38	Male	153000	0
6	61	Male	71000	1
7	27	Male	102000	0
8	38	Male	33000	1
9	42	Female	69000	0

In [4]:

heloc.shape[1]

Out[4]:

In [5]:

## That Male/ Female split is going to be a problem
## Our data needs preprocessing:

## There are lots of ways to do this!

def sex_recode(sex):
    """
    everything inside the quotes is documentation
    """
    if sex =="Female":
        return 1
    elif sex == "Male":
        return 0
    else:
        return 2

In [6]:

sex_recode("amy") ## TEst

Out[6]:

In [7]:

heloc['Sex']=heloc['Sex'].apply(sex_recode)

heloc.head(5)

Out[7]:

.dataframe tbody tr th:only-of-type {
vertical-align: middle;
}

.dataframe tbody tr th {
vertical-align: top;
}

.dataframe thead th {
text-align: right;
}

	Age	Sex	Income
0	30	1	101000
1	25	0	86000
2	20	0	50000
3	26	0	58000
4	18	1	93000

In [8]:

## Let's split the X and Y:

X = heloc[['Age', 'Sex', 'Income']]
Y = heloc[['HELOC']]

Y.tail()

Out[8]:

.dataframe tbody tr th:only-of-type {
vertical-align: middle;
}

.dataframe tbody tr th {
vertical-align: top;
}

.dataframe thead th {
text-align: right;
}

	HELOC
495	1
496	1
497	0
498	0
499	1

In [9]:

#and build a model

tree_heloc= tree.DecisionTreeClassifier(max_depth= 2, random_state = 2021)
tree_heloc= tree_heloc.fit(X, Y)

In [10]:

tree.plot_tree(tree_heloc, filled = True, feature_names = ['Age', 'Sex', 'Income']);

In [15]:

?tree.DecisionTreeClassifier

In [12]:

## Let's predict on our data!
heloc_new= pandas.read_csv("data\Heloc_unlabeled.csv") 

## fix that pesky Sex variable!
heloc_new['Sex']=heloc_new['Sex'].apply(sex_recode)

## look it over
heloc_new.head(3)

Out[12]:

.dataframe tbody tr th:only-of-type {
vertical-align: middle;
}

.dataframe tbody tr th {
vertical-align: top;
}

.dataframe thead th {
text-align: right;
}

	Age	Sex	Income
0	25	1	45000
1	23	0	22000
2	50	1	17000

In [13]:

## Trying our model, and making a new set:
Y_new = tree_heloc.predict(heloc_new)

In [14]:

## I like to make a new data.frame:
heloc_p = heloc_new

heloc_p['Heloc']=Y_new ## add in the new column of predictions

## view!
heloc_p

Out[14]:

.dataframe tbody tr th:only-of-type {
vertical-align: middle;
}

.dataframe tbody tr th {
vertical-align: top;
}

.dataframe thead th {
text-align: right;
}

	Age	Sex	Income	Heloc
0	25	1	45000	0
1	23	0	22000	0
2	50	1	17000	0
3	28	1	38000	0
4	56	0	24000	1
5	50	1	24000	0
6	22	0	27000	0
7	22	0	40000	0
8	21	0	41000	0
9	29	1	45000	0
10	64	0	47000	1
11	47	1	48000	0
12	55	1	49000	0
13	25	0	54000	1
14	42	0	113000	1
15	40	1	118000	0
16	38	0	153000	1
17	63	0	156000	1
18	51	1	172000	0
19	51	0	43000	1

License

Icon for the Creative Commons Attribution-NonCommercial 4.0 International License

Decision Trees¶

License

Share This Book