Spaces:
Runtime error
Runtime error
File size: 4,458 Bytes
fb19adb f9298b1 fb19adb 9b048cd fb19adb |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 |
import streamlit as st
import networkx as nx
import matplotlib.pyplot as plt
import pandas as pd
from sentence_transformers import SentenceTransformer
from sklearn.metrics.pairwise import cosine_similarity
from sklearn.cluster import KMeans
def main():
st.title("Financial Graph App")
st.write("Enter a financial sentence and see its similarity to predefined keywords.")
# User input
financial_sentence = st.text_area("Enter the financial sentence", value="")
# Check if the user entered a sentence
if financial_sentence.strip() != "":
# Predefined keywords
keywords = [
"Finance",
"Fiscal",
"Quarterly results",
"Revenue",
"Profit",
"Loss",
"Net income",
"Gross margin",
"Operating expenses",
"Cash flow",
"Shareholders",
"Guidance",
"Forecast",
"Market performance",
"Stock price",
"Capital expenditures",
"Acquisitions",
"Mergers",
"Debt",
"Financial ratios",
"Growth",
"Outlook",
"Visibility",
"Revenue growth",
"Market trends",
"Profitability",
"Efficiency",
"Projections",
"Performance"
]
# Load the pre-trained Sentence-Transformers model
model = SentenceTransformer("sentence-transformers/all-mpnet-base-v2")
st.set_option('deprecation.showPyplotGlobalUse', False)
# Generate word embeddings for the financial sentence and keywords
sentence_embedding = model.encode([financial_sentence])
keyword_embeddings = model.encode(keywords)
# Calculate cosine similarity between the sentence embedding and keyword embeddings
similarity_scores = cosine_similarity(sentence_embedding, keyword_embeddings)[0]
# Create a graph
G = nx.Graph()
# Add the sentence embedding as a node to the graph
G.add_node(financial_sentence, embedding=sentence_embedding[0])
# Add the keyword embeddings as nodes to the graph
for keyword, embedding, similarity in zip(keywords, keyword_embeddings, similarity_scores):
G.add_node(keyword, embedding=embedding, similarity=similarity)
# Add edges between the sentence and keywords with their similarity scores as weights
for keyword, similarity in zip(keywords, similarity_scores):
G.add_edge(financial_sentence, keyword, weight=similarity)
# Perform KNN clustering on the keyword embeddings
kmeans = KMeans(n_clusters=3)
cluster_labels = kmeans.fit_predict(keyword_embeddings)
# Add cluster labels as node attributes
for node, cluster_label in zip(G.nodes, cluster_labels):
G.nodes[node]["cluster"] = cluster_label
# Set node positions using spring layout
pos = nx.spring_layout(G)
# Get unique cluster labels
unique_clusters = set(cluster_labels)
# Assign colors to clusters
cluster_colors = ["lightblue", "lightgreen", "lightyellow"]
# Draw nodes with cluster colors
nx.draw_networkx_nodes(
G,
pos,
node_color=[cluster_colors[G.nodes[node].get("cluster", 0)] for node in G.nodes],
node_size=800,
)
# Draw edges
nx.draw_networkx_edges(G, pos, edge_color="gray", width=1, alpha=0.7)
# Draw labels
nx.draw_networkx_labels(G, pos, font_size=10, font_weight="bold")
# Draw edge labels (cosine similarity scores)
edge_labels = nx.get_edge_attributes(G, "weight")
nx.draw_networkx_edge_labels(G, pos, edge_labels=edge_labels, font_size=8)
# Set plot attributes
plt.title("Financial Context and Keywords")
plt.axis("off")
# Save the graph as an image
plt.savefig("financial_graph.png")
# Show the graph
st.pyplot()
# Save the similarity scores in a CSV file
df = pd.DataFrame({"Keyword": keywords, "Cosine Similarity": similarity_scores})
st.write("Similarity Scores:")
st.dataframe(df)
if __name__ == "__main__":
main()
|