Bubble charts are well-suited for three-dimensional information illustration:
- X-axis: The timeline (e.g., months or years).
- Y-axis: The frequency or rely of revealed articles.
- Bubble Dimension: Highlights the magnitude of stories protection for every interval.
This methodology of visualization helps us determine tendencies, outliers, or important will increase in information quantity which will point out main developments for a inventory. Utilizing Python, we will:
- Fetch information information utilizing the Monetary Modeling Prep API.
- Course of the information to combination month-to-month article counts.
- Create and improve the bubble chart with interactive options.
By the tip of this tutorial, you’ll have a script that not solely visualizes information quantity but in addition permits for intuitive exploration.
Guarantee the next Python libraries are put in in your surroundings:
- requests: For fetching information from APIs.
- pandas: For information manipulation and aggregation.
- matplotlib: For creating bubble charts.
- numpy: For numerical operations like normalization.
Set up these packages utilizing pip if obligatory:
pip set up requests pandas matplotlib numpy
We are going to use the Monetary Modeling Prep API to fetch information articles associated to a given inventory ticker. The API gives article publication dates and related metadata.
import requests
import pandas as pd# Initialize variables
ticker = "NVDA" # Substitute along with your most well-liked inventory ticker
base_url = "https://financialmodelingprep.com/api/v3/stock_news"
web page = 0
max_pages = 1500 # Variety of pages to fetch (alter as wanted)
news_data = []
api_key = "" # Substitute along with your API key
# Fetch articles from a number of pages
whereas web page < max_pages:
response = requests.get(f"{base_url}?tickers={ticker}&web page={web page}&apikey={api_key}")
if response.status_code == 200:
articles = response.json()
if not articles:
break # Cease fetching if no articles are returned
news_data.lengthen(articles)
web page += 1
else:
print(f"Did not fetch information on web page {web page}. HTTP Standing Code: {response.status_code}")
break
# Convert the fetched information right into a DataFrame
df = pd.DataFrame(news_data)
On this script, we iterate over a number of pages to gather all obtainable articles, making a DataFrame for processing. Remember to change
As soon as now we have the uncooked information, we course of it to extract the publication dates and rely the variety of articles monthly.
# Convert the 'publishedDate' column to datetime format
df['publishedDate'] = pd.to_datetime(df['publishedDate'])# Extract the 12 months and month from the publication date
df['YearMonth'] = df['publishedDate'].dt.to_period('M')
# Depend the variety of articles monthly
article_counts = df['YearMonth'].value_counts().sort_index()
This step ensures that our information is grouped and summarized by month. By utilizing the to_period('M') operate, we effectively bucket dates into month-to-month intervals.
Bubble charts present an intuitive and visually compelling solution to show the aggregated information.
import matplotlib.pyplot as plt
import numpy as np# Normalize article counts for coloration mapping
norm = plt.Normalize(article_counts.min(), article_counts.max())
colours = 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 dimension scales with article rely
c=colours, # Shade map for bubbles
alpha=0.6, # Transparency stage
cmap='viridis' # Shade map
)
# Add a coloration bar to point article quantity
cbar = plt.colorbar(scatter, ax=ax)
cbar.set_label('Variety of Articles')
# Customise the chart look
ax.set_xlabel('Yr-Month')
ax.set_ylabel('Variety of Articles')
ax.set_title(f'Information Quantity Over Time for {ticker}')
plt.xticks(rotation=45)
plt.grid(alpha=0.3)
Enhancing the chart with interactive components gives an enticing expertise for customers. By hovering over bubbles, viewers can see detailed annotations for every information level.
# Add interactivity to show annotations
def update_annotation(ind):
pos = scatter.get_offsets()[ind["ind"][0]]
annot.xy = pos
textual content = f"{article_counts.index[ind['ind'][0]]}: {article_counts.iloc[ind['ind'][0]]} articles"
annot.set_text(textual content)
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 factors",
bbox=dict(boxstyle="spherical", facecolor="white", edgecolor="grey"),
arrowprops=dict(arrowstyle="->")
)
annot.set_visible(False)
# Join hover performance to the determine
def on_hover(occasion):
if occasion.inaxes == ax:
cont, ind = scatter.comprises(occasion)
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 corporations like NVIDIA (NVDA), Palantir (PLTR), or Hims & Hers Well being (HIMS), the bubble chart reveals: