In the Finding Donors project in the preparing data section, the values of 'capital-gain' and 'capital-loss' are first log-transformed using
features_raw[skewed] = data[skewed].apply(lambda x: np.log(x + 1))
Afterwards, the values are normalized using
features_raw[numerical] = scaler.fit_transform(data[numerical])
Since the MinMaxScaler is applied on data and not (as is probably intended) on features_raw, the log-transform is only used for visualization.
Could you please clarify whether this is intended and if necessary, fix this issue.
In the Finding Donors project in the preparing data section, the values of 'capital-gain' and 'capital-loss' are first log-transformed using
features_raw[skewed] = data[skewed].apply(lambda x: np.log(x + 1))Afterwards, the values are normalized using
features_raw[numerical] = scaler.fit_transform(data[numerical])Since the MinMaxScaler is applied on data and not (as is probably intended) on features_raw, the log-transform is only used for visualization.
Could you please clarify whether this is intended and if necessary, fix this issue.