1 day ago
9
Bubble charts are well-suited for three-dimensional data representation:
- X-axis: The timeline (e.g., months or years).
- Y-axis: The frequency or count of published articles.
- Bubble Size: Highlights the magnitude of news coverage for each period.
This method of visualization helps us identify trends, outliers, or significant increases in news volume that may indicate major developments for a stock. Using Python, we can:
- Fetch news data using the Financial Modeling Prep API.
- Process the data to aggregate monthly article counts.
- Create and enhance the bubble chart with interactive features.
By the end of this tutorial, you will have a script that not only visualizes news volume but also allows for intuitive exploration.
Ensure the following Python libraries are installed in your environment:
- requests: For fetching data from APIs.
- pandas: For data manipulation and aggregation.
- matplotlib: For creating bubble charts.
- numpy: For numerical operations like normalization.
Install these packages using pip if necessary:
pip install requests pandas matplotlib numpyWe will use the Financial Modeling Prep API to fetch news articles related to a given stock ticker. The API provides article publication dates and associated metadata.
import requestsimport pandas as pd
# Initialize variables
ticker = "NVDA" # Replace with your preferred stock ticker
base_url = "https://financialmodelingprep.com/api/v3/stock_news"
page = 0
max_pages = 1500 # Number of pages to fetch (adjust as needed)
news_data = []
api_key = "<your_api_key>" # Replace with your API key
# Fetch articles from multiple pages
while page < max_pages:
response = requests.get(f"{base_url}?tickers={ticker}&page={page}&apikey={api_key}")
if response.status_code == 200:
articles = response.json()
if not articles:
break # Stop fetching if no articles are returned
news_data.extend(articles)
page += 1
else:
print(f"Failed to fetch data on page {page}. HTTP Status Code: {response.status_code}")
break
# Convert the fetched data into a DataFrame
df = pd.DataFrame(news_data)
In this script, we iterate over multiple pages to collect all available articles, creating a DataFrame for processing. Be sure to replace <your_api_key> with your actual API key.
Once we have the raw data, we process it to extract the publication dates and count the number of articles per month.
# Convert the 'publishedDate' column to datetime formatdf['publishedDate'] = pd.to_datetime(df['publishedDate'])
# Extract the year and month from the publication date
df['YearMonth'] = df['publishedDate'].dt.to_period('M')
# Count the number of articles per month
article_counts = df['YearMonth'].value_counts().sort_index()
This step ensures that our data is grouped and summarized by month. By using the to_period('M') function, we efficiently bucket dates into monthly intervals.
Bubble charts provide an intuitive and visually compelling way to display the aggregated data.
import matplotlib.pyplot as pltimport numpy as np
# Normalize article counts for color mapping
norm = plt.Normalize(article_counts.min(), article_counts.max())
colors = plt.cm.viridis(norm(article_counts.values))
# Plot the bubble chart
fig, ax = plt.subplots(figsize=(12, 8))
scatter = ax.scatter(
article_counts.index.astype(str), # Convert index to string for x-axis labels
article_counts, # Y-axis represents article counts
s=article_counts * 100, # Bubble size scales with article count
c=colors, # Color map for bubbles
alpha=0.6, # Transparency level
cmap='viridis' # Color map
)
# Add a color bar to indicate article volume
cbar = plt.colorbar(scatter, ax=ax)
cbar.set_label('Number of Articles')
# Customize the chart appearance
ax.set_xlabel('Year-Month')
ax.set_ylabel('Number of Articles')
ax.set_title(f'News Volume Over Time for {ticker}')
plt.xticks(rotation=45)
plt.grid(alpha=0.3)
Enhancing the chart with interactive elements provides an engaging experience for users. By hovering over bubbles, viewers can see detailed annotations for each data point.
# Add interactivity to display annotationsdef update_annotation(ind):
pos = scatter.get_offsets()[ind["ind"][0]]
annot.xy = pos
text = f"{article_counts.index[ind['ind'][0]]}: {article_counts.iloc[ind['ind'][0]]} articles"
annot.set_text(text)
annot.get_bbox_patch().set_facecolor('white')
annot.get_bbox_patch().set_alpha(0.6)
annot = ax.annotate(
"",
xy=(0, 0),
xytext=(15, 15),
textcoords="offset points",
bbox=dict(boxstyle="round", facecolor="white", edgecolor="gray"),
arrowprops=dict(arrowstyle="->")
)
annot.set_visible(False)
# Connect hover functionality to the figure
def on_hover(event):
if event.inaxes == ax:
cont, ind = scatter.contains(event)
if cont:
update_annotation(ind)
annot.set_visible(True)
fig.canvas.draw_idle()
else:
annot.set_visible(False)
fig.canvas.draw_idle()
fig.canvas.mpl_connect("motion_notify_event", on_hover)
When visualized for companies like NVIDIA (NVDA), Palantir (PLTR), or Hims & Hers Health (HIMS), the bubble chart reveals:
- Periods of heightened media coverage typically correlate with significant market events (e.g., product launches, earnings reports, regulatory changes).
- Sustained increases in news volume may signal heightened market attention or speculation.
- Comparing media coverage across different companies may reveal relative interest and sentiment trends.
- Media Impact: Increased news coverage often precedes stock price volatility, though this requires further analysis for confirmation.
- Pattern Detection: Visualization makes patterns in data obvious at a glance, facilitating more informed analysis.
- Interactive Features: Adding interactivity enables a richer user experience, allowing real-time exploration of the data.
This Python script serves as a starting point for integrating media sentiment analysis into your research or investment workflow.
Access the full source code on GitHub: fin_visualization/src/news_volume_visualization.py
This visualization is for informational purposes only and does not constitute investment advice. Always perform your due diligence or consult a financial professional before making investment decisions.
English (US) ·