Great questions! Let's break down what Plotly supports for these 3D visualization requests:

1. 3D Bar Charts (Fully supported) #

Plotly absolutely has built-in support for 3D bar charts using the Bar3d trace type. Here's how you can adapt your original 2D bar code to create a 3D version:

from plotly.offline import plot
from plotly.graph_objs import Bar3d, Data

# Define 3D bar data
trace1 = Bar3d(
    x=['cats', 'dogs', 'monkeys'],  # Category labels on the X-axis
    y=['Group 1'],  # Add multiple values here for multi-series 3D bars
    z=[0, 0, 0],  # Base position of bars on the Z-axis
    width=[0.5, 0.5, 0.5],  # Width of each bar
    depth=[0.5],  # Depth of each bar
    height=[20, 14, 23],  # Numerical value (height) of each bar
    marker=dict(color=['rgb(255, 100, 100)', 'rgb(100, 255, 100)', 'rgb(100, 100, 255)'])
)

data = Data([trace1])
plot(data)

You can expand this to multi-series 3D bars by adding more entries to the y array and matching height values. This type of chart is perfect for visualizing data across two categorical axes (X and Y) with numerical metrics on the Z-axis.

2. 3D Pie/Donut Charts (No native support, and not recommended) #

Plotly does not offer a native 3D pie or donut chart component—and this is by design. 3D pie charts are widely panned in data visualization because they distort slice sizes, making it much harder for viewers to accurately compare values.

If you're set on a pseudo-3D effect, you could simulate it with layered 2D pie charts using slight offsets, but this still introduces visual bias and isn't ideal. For most use cases, sticking to Plotly's standard 2D pie or donut charts is the more accurate choice. Here's a quick example of a 2D donut chart:

from plotly.offline import plot
from plotly.graph_objs import Pie, Data

trace1 = Pie(
    labels=['cats', 'dogs', 'monkeys'],
    values=[20, 14, 23],
    hole=0.4  # Adds the donut shape
)

data = Data([trace1])
plot(data)

内容的提问来源于stack exchange，提问作者Naima