hidevscommunity
commited on
Commit
•
2d6b76f
1
Parent(s):
7caf3cc
Update app.py
Browse files
app.py
CHANGED
@@ -5,13 +5,41 @@ import streamlit.components.v1 as components
|
|
5 |
from sklearn.feature_extraction.text import CountVectorizer
|
6 |
from sklearn.model_selection import train_test_split
|
7 |
import pandas as pd
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
8 |
# Separate target and feature column in X and y variable
|
9 |
df = pd.read_csv('stress.csv')
|
10 |
# X will be the features
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
11 |
X = np.array(df["text"])
|
12 |
|
13 |
# y will be the target variable
|
14 |
y = np.array(df["label"])
|
|
|
|
|
|
|
|
|
|
|
15 |
cv = CountVectorizer()
|
16 |
# Load the pickled model
|
17 |
|
@@ -40,7 +68,7 @@ def app_design():
|
|
40 |
features = text # add features according to notebook
|
41 |
|
42 |
# Make a prediction when the user clicks the "Predict" button
|
43 |
-
if st.button('Predict
|
44 |
predicted_value = model_prediction(features)
|
45 |
if predicted_value == "['Stress']":
|
46 |
st.success("Your message contains Stress")
|
|
|
5 |
from sklearn.feature_extraction.text import CountVectorizer
|
6 |
from sklearn.model_selection import train_test_split
|
7 |
import pandas as pd
|
8 |
+
import nltk
|
9 |
+
import re
|
10 |
+
nltk.download('stopwords')
|
11 |
+
stemmer = nltk.SnowballStemmer("english")
|
12 |
+
from nltk.corpus import stopwords
|
13 |
+
import string
|
14 |
+
stopword=set(stopwords.words('english'))
|
15 |
# Separate target and feature column in X and y variable
|
16 |
df = pd.read_csv('stress.csv')
|
17 |
# X will be the features
|
18 |
+
def clean(text):
|
19 |
+
text = str(text).lower()
|
20 |
+
text = re.sub('\[.*?\]', '', text)
|
21 |
+
text = re.sub('https?://\S+|www\.\S+', '', text)
|
22 |
+
text = re.sub('<.*?>+', '', text)
|
23 |
+
text = re.sub('[%s]' % re.escape(string.punctuation), '', text)
|
24 |
+
text = re.sub('\n', '', text)
|
25 |
+
text = re.sub('\w*\d\w*', '', text)
|
26 |
+
text = [word for word in text.split(' ') if word not in stopword]
|
27 |
+
text=" ".join(text)
|
28 |
+
text = [stemmer.stem(word) for word in text.split(' ')]
|
29 |
+
text=" ".join(text)
|
30 |
+
return text
|
31 |
+
|
32 |
+
|
33 |
+
df["text"] = df["text"].apply(clean)
|
34 |
X = np.array(df["text"])
|
35 |
|
36 |
# y will be the target variable
|
37 |
y = np.array(df["label"])
|
38 |
+
|
39 |
+
|
40 |
+
|
41 |
+
df["text"] = df["text"].apply(clean)
|
42 |
+
|
43 |
cv = CountVectorizer()
|
44 |
# Load the pickled model
|
45 |
|
|
|
68 |
features = text # add features according to notebook
|
69 |
|
70 |
# Make a prediction when the user clicks the "Predict" button
|
71 |
+
if st.button('Predict Stress'):
|
72 |
predicted_value = model_prediction(features)
|
73 |
if predicted_value == "['Stress']":
|
74 |
st.success("Your message contains Stress")
|