Einen 3D-Effekt auf die y-Achse eines Matplotlib-Diagramms haben ⇐ Python

[Anonymous]

Ich habe versucht, diesen Balkendiagrammstil nachzuahmen, den ich online gefunden habe. Ich konnte mehrere verschiedene Teile nachahmen, kann dieses 3D-Gefühl auf der Y-Achse nicht nachahmen. Mit Python kann ich Ideen darüber, wie ich das ermöglichen könnte? Ich freue mich, bei Bedarf eine andere Bibliothek zu verwenden (ausprobiert Seeborn, aber das gleiche Problem, kein 3D -Effekt.def create_bar_chart(
countries,
values,
date_str,
title
):

"""
Create an equity allocation chart with countries sorted by allocation percentage

Parameters
----------
- countries: list of country names
- values: list of allocation percentages (0-100)
- date_str: date to display as subtitle
- title: main chart title
"""
# ---- DATA PREPARATION ----
# Sort data by values in descending order (highest allocation first)
data = sorted(zip(countries, values, strict=False), key=lambda x: x[1], reverse=True)
sorted_countries = [item[0] for item in data]
sorted_values = [item[1] for item in data]

# Reverse the lists to get highest value at the top of the chart
# This makes the chart more intuitive to read (largest values at the top)
sorted_countries.reverse()
sorted_values.reverse()

# ---- FIGURE SETUP ----
# Create figure and axis with specified dimensions (width=12, height=6 inches)
fig, ax = plt.subplots(figsize=(12, 6))

# ---- PLOTTING THE BARS ----
# Create horizontal bars with yellow color (#FFDC00)
# zorder=3 ensures bars appear in front of grid lines
# height=0.5 controls the thickness of each bar
bars = ax.barh(sorted_countries, sorted_values, color="#FFDC00", height=0.5, zorder=3)

# ---- ADDING PERCENTAGE LABELS ----
# Add text labels showing the exact percentage at the end of each bar
for i, bar in enumerate(bars):
value = sorted_values # Get the correct value for this bar
width = bar.get_width()
label_pos = width + 1 # Position label 1 unit to the right of bar end
# Place text vertically centered on the bar with 1 decimal place precision
ax.text(label_pos, bar.get_y() + bar.get_height() / 2, f"{value:.1f}%", va="center", fontsize=9)

# ---- ADDING 3D VISUAL EFFECTS ----
# Create gray divider strip on left side of chart (3D effect)
divider_width = 2
ax.add_patch(Rectangle((-divider_width, -0.5), divider_width, len(sorted_countries), color="#E6E6E6", zorder=1))

# Add 3D effect details for each country row
for i, country in enumerate(sorted_countries):
# Right shadow edge - creates depth illusion
ax.add_patch(
Polygon(
[(0, i - 0.25), (0, i + 0.25), (-0.15, i + 0.25), (-0.15, i - 0.25)],
closed=True,
color="#AAAAAA",
zorder=2,
)
)
# Top shadow edge - enhances 3D appearance
ax.add_patch(
Polygon(
[(-divider_width, i + 0.25), (0, i + 0.25), (-0.15, i + 0.25), (-divider_width - 0.15, i + 0.25)],
closed=True,
color="#C0C0C0",
zorder=2,
)
)

# ---- GRID AND AXES STYLING ----
# Add light gray grid lines for x-axis only
# zorder=0 places grid behind all other elements
ax.grid(axis="x", color="#DDDDDD", linestyle="-", linewidth=0.5, zorder=0)
ax.set_axisbelow(True) # Ensures grid is behind other elements

# Position x-axis ticks and labels at the top of the chart
ax.xaxis.tick_top()
# Create ticks from 0% to 100% in increments of 10%
ax.set_xticks(np.arange(0, 110, 10))
ax.set_xticklabels([f"{x}%" for x in range(0, 110, 10)])

# Remove all border lines around the plot
for spine in ax.spines.values():
spine.set_visible(False)

# ---- TITLE AND SUBTITLE ----
# Add main title in bold at specific position
title_obj = plt.figtext(0.09, 0.925, title, fontsize=14, fontweight="bold")
# Add date subtitle in gray below the title
date_obj = plt.figtext(0.09, 0.875, f"as of {date_str}", fontsize=10, color="gray")

# Remove y-axis tick marks but keep the country labels
ax.tick_params(axis="y", which="both", left=False)

# Set x-axis limit to 105 to leave room for percentage labels
ax.set_xlim(0, 105)

# Add padding to left margin for better appearance
plt.subplots_adjust(left=0.09)

# Adjust overall layout to accommodate title and subtitle area
plt.tight_layout(rect=[0.0, 0.0, 1.0, 0.85])

# Return the figure and axis for further customization if needed
return fig, ax
< /code>
Offenlegung - Ich habe einen Teil des Codes geändert, aber über Python mit ChatGPT generiert. (Ich habe auch versucht, Claude zu verwenden)