Matplotlib的Deephaven插件

Matplotlib的Deephaven插件。允许在Deephaven环境中打开Matplotlib图表。任何Matplotlib图表都应该默认可查看。例如

import matplotlib.pyplot as plt

fig = plt.figure()
ax = fig.subplots()  # Create a figure containing a single axes.
ax.plot([1, 2, 3, 4], [4, 2, 6, 7])  # Plot some data on the axes.

您还可以使用TableAnimation，它允许在Deephaven Table更新时更新图表。

TableAnimation用法

TableAnimation是一个由Deephaven Table中的更新驱动的Matplotlib Animation。每当正在监听的表更新时，提供的函数将再次运行。

折线图

import matplotlib.pyplot as plt
from deephaven import time_table
from deephaven.plugin.matplotlib import TableAnimation

# Create a ticking table with the sin function
tt = time_table("PT00:00:01").update(["x=i", "y=Math.sin(x)"])

fig = plt.figure()  # Create a new figure
ax = fig.subplots()  # Add an axes to the figure
(line,) = ax.plot(
    [], []
)  # Plot a line. Start with empty data, will get updated with table updates.

# Define our update function. We only look at `data` here as the data is already stored in the format we want
def update_fig(data, update):
    line.set_data([data["x"], data["y"]])

    # Resize and scale the axes. Our data may have expanded and we don't want it to appear off screen.
    ax.relim()
    ax.autoscale_view(True, True, True)


# Create our animation. It will listen for updates on `tt` and call `update_fig` whenever there is an update
ani = TableAnimation(fig, tt, update_fig)

散点图

散点图需要的数据格式与折线图不同，因此需要以不同的方式传递数据。

import matplotlib.pyplot as plt
from deephaven import time_table
from deephaven.plugin.matplotlib import TableAnimation

tt = time_table("PT00:00:01").update(
    ["x=Math.random()", "y=Math.random()", "z=Math.random()*50"]
)

fig = plt.figure()
ax = fig.subplots()
ax.set_xlim(0, 1)
ax.set_ylim(0, 1)
scat = ax.scatter([], [])  # Provide empty data initially
scatter_offsets = []  # Store separate arrays for offsets and sizes
scatter_sizes = []


def update_fig(data, update):
    # This assumes that table is always increasing. Otherwise need to look at other
    # properties in update for creates and removed items
    added = update.added()
    for i in range(0, len(added["x"])):
        # Append new data to the sources
        scatter_offsets.append([added["x"][i], added["y"][i]])
        scatter_sizes.append(added["z"][i])

    # Update the figure
    scat.set_offsets(scatter_offsets)
    scat.set_sizes(scatter_sizes)


ani = TableAnimation(fig, tt, update_fig)

多个系列

在同一个图中可以有多种类型的系列。以下是一个示例，展示了如何驱动折线图和散点图。

import matplotlib.pyplot as plt
from deephaven import time_table
from deephaven.plugin.matplotlib import TableAnimation

tt = time_table("PT00:00:01").update(
    ["x=i", "y=Math.sin(x)", "z=Math.cos(x)", "r=Math.random()", "s=Math.random()*100"]
)

fig = plt.figure()
ax = fig.subplots()
(line1,) = ax.plot([], [])
(line2,) = ax.plot([], [])
scat = ax.scatter([], [])
scatter_offsets = []
scatter_sizes = []


def update_fig(data, update):
    line1.set_data([data["x"], data["y"]])
    line2.set_data([data["x"], data["z"]])
    added = update.added()
    for i in range(0, len(added["x"])):
        scatter_offsets.append([added["x"][i], added["r"][i]])
        scatter_sizes.append(added["s"][i])
    scat.set_offsets(scatter_offsets)
    scat.set_sizes(scatter_sizes)
    ax.relim()
    ax.autoscale_view(True, True, True)


ani = TableAnimation(fig, tt, update_fig)

构建

为了创建您的构建/开发环境（如果您已经有了一个venv，请跳过前两行）

python -m venv .venv
source .venv/bin/activate
pip install --upgrade pip setuptools
pip install build deephaven-plugin matplotlib

构建

python -m build --wheel

轮文件存储在dist/中。

在deephaven-core中测试时，请注意轮文件的存储位置（例如使用pwd）。然后，按照顶级README.md中的说明将轮文件安装到您的Deephaven环境中。