deploy_ml.py

"""This is an example file of how you can deploy your Machine Learning model
after you've trained it and saved it into pickle files.

At the end of this process you would have 2 different pickle files:
- classifier; which is the LinearSVC pipeline model
- vectorizer; which is the TfIdfVectorizer model
See: Jupyter Notebook file 'Deployment' for more information. 
"""

import pickle
import sys

sys.path.append('/Users/wolfsinem/product-tagging')
from product_tagging.tags_generator import tokenized_list

# Machine Learning - model training
from sklearn.feature_extraction.text import TfidfVectorizer
from sklearn.model_selection import train_test_split
from sklearn.preprocessing import MultiLabelBinarizer
from sklearn.svm import LinearSVC
from sklearn.multiclass import OneVsRestClassifier
from sklearn.pipeline import Pipeline
from sklearn.metrics import accuracy_score

n = 5000 # number of rows
model_df = tokenized_list()
model_df = model_df[:n]

target_variable = model_df['tags']

mlb = MultiLabelBinarizer()
target_variable = mlb.fit_transform(target_variable)


vectorizer = TfidfVectorizer(strip_accents='unicode', 
                             analyzer='word', 
                             ngram_range=(1,3), 
                             stop_words='english',
                             token_pattern=r'\w{3,}'
                            )

# fit the independent features
independent_variable = vectorizer.fit_transform(model_df['description'])

print('Independent variable shape: {}'.format(independent_variable.shape))
print('Target variable shape: {}'.format(target_variable.shape))

X_train, X_test, y_train, y_test = train_test_split(
                                        independent_variable, 
                                        target_variable, 
                                        test_size=0.2, 
                                        random_state=42, 
                                        )

Linear_pipeline = Pipeline([
                ('clf', OneVsRestClassifier(LinearSVC(
                                                class_weight='balanced',
                                                random_state=42,
                                                tol=1e-1,
                                                C=8.385), 
                                                n_jobs=-1)),
            ])

Linear_pipeline.fit(X_train, y_train)

prediction = Linear_pipeline.predict(X_test)
print('Accuracy for LinearSVC is {}'.format(accuracy_score(y_test, prediction)))

# To use the machine learning for deployement you need to save the ML model
# as a pickle file first. Take into consideration that it's a really big file
# up to 5GB
with open('classifier2', 'wb') as picklefile:
    pickle.dump(Linear_pipeline,picklefile)

with open('vect2', 'wb') as picklefile:
    pickle.dump(vectorizer,picklefile)

# After saving the ML model you can open the pickle file and start using the 
# ML model
with open('/Users/wolfsinem/product-tagging/data/classifier2', 'rb') as training_model:
    model = pickle.load(training_model)

with open('/Users/wolfsinem/product-tagging/data/vect2', 'rb') as training_model:
    vectorizer = pickle.load(training_model)

# This would be an example of how you can use the saved model. 
user_input_string = ["""Key Features of Alisha Solid Womens Cycling Shorts Cotton 
                    'Lycra Navy, Red, Navy,Specifications of Alisha Solid Womens 
                    Cycling Shorts Shorts Details Number of Contents in Sales 
                    Package Pack of 3 Fabric Cotton Lycra Type Cycling Shorts 
                    General Details Pattern Solid Ideal For Womens Fabric Care 
                    Gentle Machine Wash in Lukewarm Water, Do Not Bleach Additional 
                    Details Style Code ALTHT_3P_21 In the Box 3 shorts"""]


user_input_string = vectorizer.transform(user_input_string).toarray()
label = model.predict(user_input_string)
tags = mlb.inverse_transform(label)
print(tags)

# The output of the model are the following tags:
#  [('alisha', 'care', 'cotton', 'cycling', 'gentle', 'ideal', 'lukewarm', 
# 'lycra', 'machine', 'navy', 'pack', 'package', 'pattern', 'red', 'shorts', 
# 'solid', 'wash', 'water', 'women')]