Update app.py
Browse files
app.py
CHANGED
@@ -19,6 +19,9 @@ from pinecone import Pinecone
|
|
19 |
import logging
|
20 |
import re
|
21 |
|
|
|
|
|
|
|
22 |
from langchain_openai import ChatOpenAI
|
23 |
from langchain_openai import OpenAIEmbeddings
|
24 |
llm = ChatOpenAI(model="gpt-4o")
|
@@ -129,56 +132,6 @@ def clean_retrieved_context(raw_context):
|
|
129 |
# Return explicitly cleaned context
|
130 |
return cleaned.strip()
|
131 |
|
132 |
-
|
133 |
-
def extract_keywords_with_gpt(context, max_tokens=100, temperature=0.3):
|
134 |
-
|
135 |
-
keyword_prompt = f"Extract 3 most important scientific keywords from the following user query:\n\n{context}"
|
136 |
-
|
137 |
-
response = client.chat.completions.create(
|
138 |
-
model="gpt-4",
|
139 |
-
messages=[
|
140 |
-
{"role": "system", "content": "You are an expert in identifying key scientific terms and concepts."},
|
141 |
-
{"role": "user", "content": keyword_prompt}
|
142 |
-
],
|
143 |
-
max_tokens=max_tokens,
|
144 |
-
temperature=temperature
|
145 |
-
)
|
146 |
-
|
147 |
-
extracted_keywords = response.choices[0].message.content.strip()
|
148 |
-
|
149 |
-
cleaned_keywords = re.sub(r'\d+\.\s*', '', extracted_keywords)
|
150 |
-
|
151 |
-
keywords_list = [kw.strip() for kw in cleaned_keywords.split("\n") if kw.strip()]
|
152 |
-
|
153 |
-
return keywords_list
|
154 |
-
|
155 |
-
def fetch_nasa_ads_references(ads_query):
|
156 |
-
"""Fetch relevant NASA ADS papers and format them for readability."""
|
157 |
-
try:
|
158 |
-
# Query NASA ADS for relevant papers
|
159 |
-
papers = ADS.query_simple(ads_query)
|
160 |
-
|
161 |
-
if not papers or len(papers) == 0:
|
162 |
-
return [("No results found", "N/A", "N/A", "N/A", "N/A", "N/A")]
|
163 |
-
|
164 |
-
# Include authors in the references
|
165 |
-
references = []
|
166 |
-
for paper in papers[:5]: # Limit to 5 references
|
167 |
-
title = paper.get('title', ['Title not available'])[0]
|
168 |
-
abstract = paper.get('abstract', 'Abstract not available')
|
169 |
-
authors = ", ".join(paper.get('author', [])[:3]) + (" et al." if len(paper.get('author', [])) > 3 else "")
|
170 |
-
bibcode = paper.get('bibcode', 'N/A')
|
171 |
-
pub = paper.get('pub', 'Unknown Journal')
|
172 |
-
pubdate = paper.get('pubdate', 'Unknown Date')
|
173 |
-
|
174 |
-
references.append((title, abstract, authors, bibcode, pub, pubdate))
|
175 |
-
|
176 |
-
return references
|
177 |
-
|
178 |
-
except Exception as e:
|
179 |
-
logging.error(f"Error fetching ADS references: {str(e)}")
|
180 |
-
return [("Error fetching references", "See logs for details", "N/A", "N/A", "N/A", "N/A")]
|
181 |
-
|
182 |
def fetch_exoplanet_data():
|
183 |
# Connect to NASA Exoplanet Archive TAP Service
|
184 |
tap_service = vo.dal.TAPService("https://exoplanetarchive.ipac.caltech.edu/TAP")
|
|
|
19 |
import logging
|
20 |
import re
|
21 |
|
22 |
+
from utils.ads_references import fetch_nasa_ads_references, extract_keywords_with_gpt
|
23 |
+
|
24 |
+
|
25 |
from langchain_openai import ChatOpenAI
|
26 |
from langchain_openai import OpenAIEmbeddings
|
27 |
llm = ChatOpenAI(model="gpt-4o")
|
|
|
132 |
# Return explicitly cleaned context
|
133 |
return cleaned.strip()
|
134 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
135 |
def fetch_exoplanet_data():
|
136 |
# Connect to NASA Exoplanet Archive TAP Service
|
137 |
tap_service = vo.dal.TAPService("https://exoplanetarchive.ipac.caltech.edu/TAP")
|