Whats the purpose of encoding if I cannot use it in predict

Question

This is a follow-up to this question.

I thought the reason we did OneHotEncoding was to convert string data to a numpy array right?

Then, the Predict statement val_predictions = soccer_model.predict(val_X) should work as it is using encoded data.

Here is the code I had so far:

import numpy as np
import pandas as pd
from sklearn.metrics import mean_absolute_error
from sklearn.model_selection import train_test_split
from sklearn.preprocessing import OneHotEncoder
from sklearn.tree import DecisionTreeRegressor

# Set option to display all the rows and columns in the dataset. If there are more rows, adjust number accordingly.
pd.set_option('display.max_rows', 5000)
pd.set_option('display.max_columns', 500)
pd.set_option('display.width', 1000)

# Pandas needs you to define the column as date before its imported and then call the column and define as a date
# hence this step.
date_col = ['Date']
df = pd.read_csv(
    r'C:\Users\harsh\Documents\My Dream\Desktop\Machine Learning\Attempt1\Historical Data\Concat_Cleaned.csv'
    , parse_dates=date_col, skiprows=0, low_memory=False)

# Converting/defining the columns
# Before you define column types, you need to fill all NaN with a value. We will be reconverting them later
df = df.fillna(101)
# Defining column types
convert_dict = {'League_Division': str,
                'HomeTeam': str,
                'AwayTeam': str,
                'Full_Time_Home_Goals': int,
                'Full_Time_Away_Goals': int,
                'Full_Time_Result': str,
                'Half_Time_Home_Goals': int,
                'Half_Time_Away_Goals': int,
                'Half_Time_Result': str,
                'Attendance': int,
                'Referee': str,
                'Home_Team_Shots': int,
                'Away_Team_Shots': int,
                'Home_Team_Shots_on_Target': int,
                'Away_Team_Shots_on_Target': int,
                'Home_Team_Hit_Woodwork': int,
                'Away_Team_Hit_Woodwork': int,
                'Home_Team_Corners': int,
                'Away_Team_Corners': int,
                'Home_Team_Fouls': int,
                'Away_Team_Fouls': int,
                'Home_Offsides': int,
                'Away_Offsides': int,
                'Home_Team_Yellow_Cards': int,
                'Away_Team_Yellow_Cards': int,
                'Home_Team_Red_Cards': int,
                'Away_Team_Red_Cards': int,
                'Home_Team_Bookings_Points': float,
                'Away_Team_Bookings_Points': float,
                }

df = df.astype(convert_dict)

# Reverting the replace values step to get original dataframe and with the defined filetypes
df = df.replace('101', np.NAN, regex=True)
df = df.replace(101, np.NAN, regex=True)

# Clean dataset by dropping null rows
data = df.dropna(axis=0)

# Column that you want to predict = y
y = data.Full_Time_Home_Goals

# Columns that are inputted into the model to make predictions (dependants), Cannot be column y
features = ['HomeTeam', 'AwayTeam', 'Full_Time_Away_Goals', 'Full_Time_Result']
# Create X
X = data[features]

# Split into validation and training data
train_X, val_X, train_y, val_y = train_test_split(X, y, random_state=1)

# Specify Model
soccer_model = DecisionTreeRegressor(random_state=1)

# Define and train OneHotEncoder to transform numerical data to a numeric array
enc = OneHotEncoder(handle_unknown='ignore')
enc.fit(train_X)

transformed_train_X = enc.transform(train_X)

# Fit Model
soccer_model.fit(transformed_train_X, train_y)

#  Make validation predictions and calculate mean absolute error
val_predictions = soccer_model.predict(val_X)
val_mae = mean_absolute_error(val_predictions, val_y)
print("Validation MAE when not specifying max_leaf_nodes : {:,.0f}".format(val_mae))

The error I am getting is at

val_predictions = soccer_model.predict(val_X)

The error That I get is:

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

You can find my sample dataset here

Answer 1

Look at these again:

transformed_train_X = enc.transform(train_X)

# Fit Model
soccer_model.fit(transformed_train_X, train_y)

What you did was to do an encoding to train_X and use the encoded data to fit a model soccer_model. That's what the model would expect. So to use it, you should apply the same encoding, that is, you should do this:

transformed_val_X = enc.transform(val_X)

#  Make validation predictions and calculate mean absolute error
val_predictions = soccer_model.predict(transformed_val_X)