How to encode text to numeric using fit_transform

 Hi all,

Qn) How to encode text to numeric using fit_transform

Ans)

import numpy as np

import pandas as pd

from sklearn.feature_extraction.text import CountVectorizer

df = pd.read_csv("hupassg.tsv", sep='\t', encoding='ISO-8859-1');

vectorizer1 = CountVectorizer(max_features = 10000, ngram_range=(1, 3), stop_words='english')

count_vector1 = vectorizer1.fit_transform(df['clean_assg'])

feature_names1 = vectorizer1.get_feature_names_out()

data1 = df[['assg_set','clean_assg','final_score']].copy()

X = count_vectors1.toarray()

y = data1['final_score'].to_numpy()

X_train, X_test, y_train, y_test = train_test_split(X, y, test_size = 0.3)

Save and load model in python using pickle

 Dear all,

Qn) How to save and load a model using Pickle in Python

Ans)

import pickle

filename = 'hupscoringsvm.sav'

pickle.dump(model, open(filename, 'wb'))

--------

import pickle

filename = 'hupscoringsvm.sav'

loaded_model = pickle.load(open(filename, 'rb'))

y_pred=loaded_model.predict(X_test)

print(y_pred)

Save Feature using Pickle

 Dear all,

Qn. How can we save features using pickle in python

Ans. 

import pickle

pickle.dump(X, open('X.pkl', 'wb'))

pickle.dump(y, open('y.pkl', 'wb'))

X = pickle.load(open('X.pkl', 'rb'))

y = pickle.load(open('y.pkl', 'rb'))

//And split to make testing

X_train, X_test, y_train, y_test = train_test_split(X, y, test_size = 0.3)