fixed model errors and edited code to fix nltk error, still need HN_100k csv

Author: Witterone

Pipfile

@@ -16,6 +16,10 @@ fastapi = "*"
 pandas = "*"
 uvicorn = "*"
 plotly = "*"
+nltk = "*"
+re = "*"
+vaderSentiment = "*"
+warnings = "*"
 
 [requires]
 python_version = "3.7"

app/api/__pycache__/model.cpython-37.pyc

@@ -1,10 +1,3 @@
-# -*- coding: utf-8 -*-
-"""
-Created on Wed Jul 29 17:07:08 2020
-
-@author: Ronin
-"""
-
 from warnings import filterwarnings
 filterwarnings("ignore")
 
@@ -16,6 +9,9 @@
 import urllib
 import re
 import pandas as pd
+import os
+
+
 
 
 from vaderSentiment.vaderSentiment import SentimentIntensityAnalyzer
@@ -24,9 +20,9 @@
 lemmatizer = WordNetLemmatizer()
 stemmer = PorterStemmer()
 
-#If you get stopwords error please uncomment the following two lines.
-# nltk.download('stopwords')
-# nltk.download('wordnet')
+#If you get stopwords error pleasew uncomment the following two lines.
+nltk.download('stopwords')
+nltk.download('wordnet')
 
 def get_compund_score(text):
         score = analyzer.polarity_scores(text)
@@ -96,8 +92,21 @@ def get_cummulative_score_for_user(username):
   return score
 
 
+def get_scores_by_user():
+  dir = os.getcwd()
+  file_name = 'hn_sentiments100k.csv'
+  df = pd.read_csv(os.path.normcase(os.path.join(dir, file_name)))
+  df1 = df.groupby(['by']).agg({'clean_vader_score': "sum"})
+  df2 = df1.sort_values(['clean_vader_score'], ascending=True)
 
+  return df2
+
+    
 def main():
+
+    out = get_scores_by_user()
+    print (out.head(10))
+
     id = 23970146
     print(f"Score for the comment id {id} is: ",  get_score_by_comment_id(id))
 
@@ -114,4 +123,4 @@ def main():
 
 
 if __name__ == "__main__":
-    main()
+    main()

app/api/__pycache__/predict.cpython-37.pyc

@@ -6,7 +6,7 @@
 
 from pydantic import BaseModel, Field, validator
 
-import model
+from .model import *
 
 log = logging.getLogger(__name__)
 
@@ -19,8 +19,8 @@ class Item(BaseModel):
 
 
     comment_id : int = Field(..., example=23970146)
-    user_name : str = Field(..., example='gmfawcett')
-    
+    #user_name : str = Field(..., example='gmfawcett')
+    #comment_id_list : list = Field(..., example=[23970146,457634])
 
     def to_df(self):
         """Convert pydantic object to pandas dataframe with 1 row."""
@@ -41,10 +41,12 @@ async def predict(item: Item):
     X_new = item.to_df()
     log.info(X_new)
 
-    y_pred = model.get_score_by_comment_id(item.comment_id)
-    U_pred = model.get_cummulative_score_for_user(item.user_name)
+    y_pred = get_score_by_comment_id(item.comment_id)
+    #U_pred = get_cummulative_score_for_user(item.user_name)
+    l_pred = get_scores_by_user()
     return {
 
         'Score for comment from id': y_pred,
-        'user cumulative comment score': U_pred
+        #'user cumulative comment score': U_pred,
+        'Score for list of comment ids': l_pred
     }

app/api/model.py

@@ -2,3 +2,10 @@ urllib
 json
 pandas
 logging
+fastapi
+pandas
+plotly
+nltk
+re
+vaderSentiment
+warnings