Update app.py
Browse files
app.py
CHANGED
@@ -3,9 +3,9 @@ import pandas as pd
|
|
3 |
import os
|
4 |
from pandasai import SmartDataframe
|
5 |
from pandasai.llm import OpenAI
|
6 |
-
from dotenv import load_dotenv
|
7 |
import tempfile
|
8 |
import matplotlib.pyplot as plt
|
|
|
9 |
|
10 |
# Load environment variables
|
11 |
openai_api_key = os.getenv("OPENAI_API_KEY")
|
@@ -18,30 +18,83 @@ if not openai_api_key:
|
|
18 |
# Initialize the LLM
|
19 |
llm = OpenAI(api_token=openai_api_key)
|
20 |
|
21 |
-
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
22 |
|
23 |
# Instructions
|
24 |
with st.sidebar:
|
25 |
-
st.header("Instructions")
|
26 |
st.markdown(
|
27 |
-
"1.
|
28 |
-
"2.
|
29 |
-
"3. Enter a
|
|
|
|
|
|
|
|
|
30 |
)
|
31 |
|
32 |
-
|
|
|
33 |
|
34 |
-
if
|
35 |
-
|
36 |
-
df = pd.read_csv(uploaded_file)
|
37 |
st.write("### Data Preview")
|
38 |
st.dataframe(df.head(10))
|
39 |
|
40 |
# Create SmartDataFrame
|
41 |
chat_df = SmartDataframe(df, config={"llm": llm})
|
42 |
|
43 |
-
st.write("### Chat with Your Data")
|
44 |
-
user_query = st.text_input("Enter your question about the data:")
|
45 |
|
46 |
if user_query:
|
47 |
try:
|
@@ -51,7 +104,7 @@ if uploaded_file:
|
|
51 |
st.error(f"Error: {e}")
|
52 |
|
53 |
st.write("### Generate and View Graphs")
|
54 |
-
plot_query = st.text_input("Enter a query to generate a graph (e.g., '
|
55 |
|
56 |
if plot_query:
|
57 |
try:
|
|
|
3 |
import os
|
4 |
from pandasai import SmartDataframe
|
5 |
from pandasai.llm import OpenAI
|
|
|
6 |
import tempfile
|
7 |
import matplotlib.pyplot as plt
|
8 |
+
from datasets import load_dataset
|
9 |
|
10 |
# Load environment variables
|
11 |
openai_api_key = os.getenv("OPENAI_API_KEY")
|
|
|
18 |
# Initialize the LLM
|
19 |
llm = OpenAI(api_token=openai_api_key)
|
20 |
|
21 |
+
def validate_and_clean_dataset(dataframe):
|
22 |
+
# Placeholder for dataset validation and cleaning logic
|
23 |
+
return dataframe
|
24 |
+
|
25 |
+
def load_dataset_into_session():
|
26 |
+
input_option = st.radio(
|
27 |
+
"Select Dataset Input:",
|
28 |
+
["Use Repo Directory Dataset", "Use Hugging Face Dataset", "Upload CSV File"],
|
29 |
+
)
|
30 |
+
|
31 |
+
# Option 1: Load dataset from the repo directory
|
32 |
+
if input_option == "Use Repo Directory Dataset":
|
33 |
+
file_path = "./source/test.csv"
|
34 |
+
if st.button("Load Dataset"):
|
35 |
+
try:
|
36 |
+
st.session_state.df = pd.read_csv(file_path)
|
37 |
+
st.session_state.df = validate_and_clean_dataset(st.session_state.df)
|
38 |
+
st.success(f"File loaded successfully from '{file_path}'!")
|
39 |
+
except Exception as e:
|
40 |
+
st.error(f"Error loading dataset from the repo directory: {e}")
|
41 |
+
|
42 |
+
# Option 2: Load dataset from Hugging Face
|
43 |
+
elif input_option == "Use Hugging Face Dataset":
|
44 |
+
dataset_name = st.text_input(
|
45 |
+
"Enter Hugging Face Dataset Name:", value="HUPD/hupd"
|
46 |
+
)
|
47 |
+
if st.button("Load Hugging Face Dataset"):
|
48 |
+
try:
|
49 |
+
dataset = load_dataset(dataset_name, split="train", trust_remote_code=True)
|
50 |
+
if hasattr(dataset, "to_pandas"):
|
51 |
+
st.session_state.df = dataset.to_pandas()
|
52 |
+
else:
|
53 |
+
st.session_state.df = pd.DataFrame(dataset)
|
54 |
+
st.session_state.df = validate_and_clean_dataset(st.session_state.df)
|
55 |
+
st.success(f"Hugging Face Dataset '{dataset_name}' loaded successfully!")
|
56 |
+
except Exception as e:
|
57 |
+
st.error(f"Error loading Hugging Face dataset: {e}")
|
58 |
+
|
59 |
+
# Option 3: Upload CSV File
|
60 |
+
elif input_option == "Upload CSV File":
|
61 |
+
uploaded_file = st.file_uploader("Upload a CSV File:", type=["csv"])
|
62 |
+
if uploaded_file:
|
63 |
+
try:
|
64 |
+
st.session_state.df = pd.read_csv(uploaded_file)
|
65 |
+
st.session_state.df = validate_and_clean_dataset(st.session_state.df)
|
66 |
+
st.success("File uploaded successfully!")
|
67 |
+
except Exception as e:
|
68 |
+
st.error(f"Error reading uploaded file: {e}")
|
69 |
+
|
70 |
+
st.title("Chat with Patent Dataset Using PandasAI")
|
71 |
|
72 |
# Instructions
|
73 |
with st.sidebar:
|
74 |
+
st.header("Instructions:")
|
75 |
st.markdown(
|
76 |
+
"1. Select how you want to input the dataset.\n"
|
77 |
+
"2. Upload, select, or fetch the dataset using the provided options.\n"
|
78 |
+
"3. Enter a question to interact with the patent data.\n"
|
79 |
+
" - Example: 'Predict if the patent will be accepted.'\n"
|
80 |
+
" - Example: 'What is the primary classification of this patent?'\n"
|
81 |
+
" - Example: 'Summarize the abstract of this patent.'\n"
|
82 |
+
"4. Enter a query to generate and view graphs based on patent attributes.\n"
|
83 |
)
|
84 |
|
85 |
+
# Load dataset into session
|
86 |
+
load_dataset_into_session()
|
87 |
|
88 |
+
if "df" in st.session_state:
|
89 |
+
df = st.session_state.df
|
|
|
90 |
st.write("### Data Preview")
|
91 |
st.dataframe(df.head(10))
|
92 |
|
93 |
# Create SmartDataFrame
|
94 |
chat_df = SmartDataframe(df, config={"llm": llm})
|
95 |
|
96 |
+
st.write("### Chat with Your Patent Data")
|
97 |
+
user_query = st.text_input("Enter your question about the patent data (e.g., 'Predict if the patent will be accepted.'):")
|
98 |
|
99 |
if user_query:
|
100 |
try:
|
|
|
104 |
st.error(f"Error: {e}")
|
105 |
|
106 |
st.write("### Generate and View Graphs")
|
107 |
+
plot_query = st.text_input("Enter a query to generate a graph (e.g., 'Plot the number of patents by filing year.'):")
|
108 |
|
109 |
if plot_query:
|
110 |
try:
|