Fit_transform method for scaling data throws value error. Please help me resolve it

Question

I am learning Machine Learning from scratch from a book. I am sorry if this is a naive question or something that was discussed already here. I reviewed various other similar posts here and learned that I need to use Label Encoder to resolve this but I am not sure how to code Label Encoder and hoping someone here will help me. I really appreciate your time and your help with this.

Code:

housing_feature_engineered = pd.read_csv("todaytest.csv")
from sklearn.preprocessing import StandardScaler
scaler = StandardScaler()
housing_scaled = scaler.fit_transform(housing_feature_engineered)
housing_scaled

Output:

ValueError: could not convert string to float: 'INLAND'

Answer 1

You can only input numeric values to the StandardScaler function. You are getting this error because the data contains string expressions.

Solution to the problem:

• have an example dataframe -->>> df_example

• You want to scale the ex_column_name column in the dataframe. Now you need to eliminate the string expressions contained in this df.

• solution based on the example you gave (here we're deleting the lines that contain "INLAND"):

  df_example=df_example[df_example["ex_column_name"]!='INLAND']

now you can scale your data :)