Pdp.pdp_isolate function requires a new positional argument 'features'

Intro to Machine Learning (2018)
1

Hello,

I am trying to pdp_isolate the ‘Sex’ feature in the Titanic competition, but I get an error if I try running the same code as shown in the lesson. Contrary to the original lesson’s code, pdp_isolate requires a new argument in addition to feat: features (the features of the model). So I did:

features = [feat for feat in df_raw.columns]

['PassengerId',
 'Survived',
 'Pclass',
 'Name',
 'Sex',
 'Age',
 'SibSp',
 'Parch',
 'Ticket',
 'Fare',
 'Cabin',
 'Embarked']

and then:

def plot_pdp(feat, clusters=None, feat_name=None):
    feat_name = feat_name or feat
    p = pdp.pdp_isolate(m, df_raw, features, feat)
    return plot_pdp(p, feat_name, plot_lines=True, cluster=clusters is not None, n_cluster_centers=clusters)
plot_pdp('Sex')

But I get this error:

TypeError: can't multiply sequence by non-int of type 'float'

Did anyone have any trouble using this function? What does this TypeError mean?

2

See the discussion here, which mentions the fix:

def plot_pdp(feat, clusters=None, feat_name=None):
    feat_name = feat_name or feat
    p = pdp.pdp_isolate(model=m, dataset=x, model_features=x.columns, feature=feat)
    return pdp.pdp_plot(p, feat_name, plot_lines=True,
                        cluster=clusters is not None,
                        n_cluster_centers=clusters)