def feedback(pipeline,parameters,ultimateTraindf):
# Read feedback data
fbkdf = pd.read_csv("../../results/feedback.csv")
# Extract incorrect predictions
fbkdf = fbkdf.loc[fbkdf['check'] == False]
# Combine with rest of the training data
fbkdf = fbkdf.append(ultimateTraindf, ignore_index=True)
# Create matrix for learning and prediction
X, y = fbkdf['properIngredients'], fbkdf['cuisine'].as_matrix()
# Split further into training and validation sets
Xtrain, Xvalidate, ytrain, yValidate = train_test_split(X, y, train_size=0.7)
# Initialize gridSearchClassifierCV Classifier with parameters
gridSearchClassifier = cookut.getGridSearchCv(pipeline, parameters)
# Fit the gridSearchClassifier on Training Set
gridSearchClassifier.fit(Xtrain,ytrain)
# Calculate best set of parameters and make predictions on validation set
return validate(parameters, gridSearchClassifier, Xvalidate, yValidate)
def feedback(pipeline, parameters, ultimateTraindf):
# Read feedback data
fbkdf = pd.read_csv("../../results/feedback.csv")
# Extract incorrect predictions
fbkdf = fbkdf.loc[fbkdf['check'] == False]
# Combine with rest of the training data
fbkdf = fbkdf.append(ultimateTraindf, ignore_index=True)
# Create matrix for learning and prediction
X, y = fbkdf['properIngredients'], fbkdf['cuisine'].as_matrix()
# Split further into training and validation sets
Xtrain, Xvalidate, ytrain, yValidate = train_test_split(X,
y,
train_size=0.7)
# Initialize gridSearchClassifierCV Classifier with parameters
gridSearchClassifier = cookut.getGridSearchCv(pipeline, parameters)
# Fit the gridSearchClassifier on Training Set
gridSearchClassifier.fit(Xtrain, ytrain)
# Calculate best set of parameters and make predictions on validation set
return validate(parameters, gridSearchClassifier, Xvalidate, yValidate)
def getBestParameters(pipeline,parameters):
# Load training data
traindf = pd.read_json('../../data/train.json')
# Remove everything but alphabets and then Lemmatize. Also remove extra whitespace
traindf['properIngredients'] = [' '.join([WordNetLemmatizer().lemmatize(re.sub('[^A-Za-z]', ' ', line)) for line in lists]).strip() for lists in traindf['ingredients']]
# Create learning matrix
X, y = traindf['properIngredients'], traindf['cuisine'].as_matrix()
# Split into Training and Validation Sets
Xtrain, Xvalidate, ytrain, yValidate = train_test_split(X, y, train_size=0.7)
# Initialize gridSearchClassifierCV Classifier with parameters
gridSearchClassifier = cookut.getGridSearchCv(pipeline, parameters)
# Fit/train the gridSearchClassifier on Training Set
gridSearchClassifier.fit(Xtrain, ytrain)
# Make predictions on validation set and calculate best set of parameters
bestParameters,predictions=cookVal.validate(parameters, gridSearchClassifier, Xvalidate, yValidate)
# Initialize DataFrame for feedback loop
valdf = pd.DataFrame(index = Xvalidate.index.values)
# Add ingredients column
valdf=valdf.join(Xvalidate)
