📝 Exercise M1.04#
The goal of this exercise is to evaluate the impact of using an arbitrary integer encoding for categorical variables along with a linear classification model such as Logistic Regression.
To do so, let’s try to use
OrdinalEncoder to preprocess the categorical
variables. This preprocessor is assembled in a pipeline with
LogisticRegression. The generalization performance of the pipeline can be
evaluated by cross-validation and then compared to the score obtained when
OneHotEncoder or to some other baseline score.
First, we load the dataset.
import pandas as pd adult_census = pd.read_csv("../datasets/adult-census.csv")
target_name = "class" target = adult_census[target_name] data = adult_census.drop(columns=[target_name, "education-num"])
In the previous notebook, we used
automatically select columns with a specific data type (also called
Here, we will use this selector to get only the columns containing strings
object dtype) that correspond to categorical features in our
from sklearn.compose import make_column_selector as selector categorical_columns_selector = selector(dtype_include=object) categorical_columns = categorical_columns_selector(data) data_categorical = data[categorical_columns]
Define a scikit-learn pipeline composed of an
OrdinalEncoder and a
OrdinalEncoder can raise errors if it sees an unknown category at
prediction time, you can set the
unknown_value parameters. You can refer to the scikit-learn
for more details regarding these parameters.
from sklearn.pipeline import make_pipeline from sklearn.preprocessing import OrdinalEncoder from sklearn.linear_model import LogisticRegression # Write your code here.
Your model is now defined. Evaluate it using a cross-validation using
Be aware that if an error happened during the cross-validation,
cross_validate will raise a warning and return NaN (Not a Number) as scores.
To make it raise a standard Python exception with a traceback, you can pass
error_score="raise" argument in the call to
exception will be raised instead of a warning at the first encountered problem
cross_validate will stop right away instead of returning NaN values.
This is particularly handy when developing complex machine learning pipelines.
from sklearn.model_selection import cross_validate # Write your code here.
Now, we would like to compare the generalization performance of our previous
model with a new model where instead of using an
OrdinalEncoder, we will use
OneHotEncoder. Repeat the model evaluation using cross-validation. Compare
the score of both models and conclude on the impact of choosing a specific
encoding strategy when using a linear model.
from sklearn.preprocessing import OneHotEncoder # Write your code here.