Extracting features from a Pandas Dataframe does not work out of the box with DictVectorizer. This is an efficient way to extract your categorical features

DictVectorizer_Pandas.py

from sklearn.feature_extraction import DictVectorizer
import pandas as pd

df = pd.DataFrame({"user_name": ["a", "b", "c"]})
fe_lm = DictVectorizer()
design_lm = fe_lm.fit_transform(df.to_dict(orient="records"))

# Note that this solution is *MUCH* faster (60 times) than transposing and converting into a dictionary:
# http://fastml.com/converting-categorical-data-into-numbers-with-pandas-and-scikit-learn/ is much slower