条件线条颜色会导致 ValueError

Question

我正在尝试使用 matplotlib.pytplot 绘制线图，并拥有形状为

(42,7)

的数据框 df 。数据框具有以下结构（仅显示相关列）：

timepoint   value   point
2021-01-01   10      0
2021-02-01   20      0
....
2021-11-01   10      0
2021-12-01   50      1
2022-01-01   60      1
...

我尝试按以下方式绘制条件颜色（每个点 = 0 的值是蓝色，每个点 = 1 的值是红色）：

import numpy as np
col = np.where(df['point'] == 0, 'b', 'r')


plt.plot(df['timepoint'], df['value'], c=col)
plt.show()

我收到错误消息：

值错误：数组（['b'，'b'，'b'，'b'，'b'，'b'，'b'，'b'，'b'，'b'， 'b', 'b', 'b', 'b'、'r'、'r'、'r'、'r'、'r'、'r'、'r'、'r'、'r'、'r'、'r'、'r ', 'r', 'r', 'r', 'r', 'r', 'r', 'r', 'r', 'r', 'r', 'r', 'r', 'r ', 'r', 'r', 'r'], dtype='

当我查看这个问题 ValueError: Invalid RGBA argument: What is Cause this error? 时，我找不到任何解决方案，因为我的颜色数组的形状是：

col.shape

是

(42, )

Answer 1

我认为您不能在此处使用

np.where

来定义折线图的颜色（就像他们在 here 中使用散点图那样），因为在前者中，

c

参数需要单个整条线的颜色，而在后者中，您可以提供将映射到点的颜色数组。

因此，这是解决这个问题的一种可能的选择：

(_, p0), (_, others) = df.groupby(df["point"].eq(0), sort=False)

plt.plot(p0.timepoint, p0.value, c="b", label="0 Points")
plt.plot(others.timepoint, others.value, c="r", label="Others")

plt.legend()
plt.show();

或与

pivot

/

plot

:

other_points = df["point"].loc[lambda s: s.ne(0)].unique()

(
    df.pivot(index="timepoint", columns="point", values="value")
        .plot(style={0: "b", **{k: "r" for k in other_points}})
)

输出：

Answer 2

您遇到的错误是因为 plt.plot() 函数中的 c 参数需要每个数据点的有效颜色规范，但您提供了一个颜色名称数组（'b' 表示蓝色，'r'红色）这不是 c 参数的有效值。

要为折线图绘制条件颜色，您可以使用循环迭代 DataFrame 行并使用所需的颜色绘制每个段。以下是如何做到这一点（使用示例数据框）：

import pandas as pd
import matplotlib.pyplot as plt

# Sample DataFrame
data = {
    'timepoint': ['2021-01-01', '2021-02-01', '2021-11-01', '2021-12-01', '2022-01-01'],
    'value': [10, 20, 10, 50, 60],
    'point': [0, 0, 0, 1, 1]
}

df = pd.DataFrame(data)

# Convert 'timepoint' column to datetime type
df['timepoint'] = pd.to_datetime(df['timepoint'])

# Initialize variables to keep track of the previous point value and color
prev_point = None
prev_color = None

fig, ax = plt.subplots()

for index, row in df.iterrows():
    # Check the 'point' value for the current row
    current_point = row['point']

    # Set the color based on the 'point' value
    if current_point == 0:
        color = 'b'  # Blue for point=0
    else:
        color = 'r'  # Red for point=1

    # If the point value has changed, create a new segment in the plot
    if prev_point is not None and current_point != prev_point:
        ax.plot(df.loc[df.index[index - 1]:index - 1, 'timepoint'], df.loc[df.index[index - 1]:index - 1, 'value'], c=prev_color, label=f'Point {prev_point}', marker='o')  # Add marker='o' to display points

    ax.plot([row['timepoint']], [row['value']], c=color, marker='o')  # Add marker='o' to display points

    # Update previous point and color
    prev_point = current_point
    prev_color = color

# Add labels, legend, and show the plot
ax.set_xlabel('Timepoint')
ax.set_ylabel('Value')
ax.legend()
plt.xticks(rotation=45)  # Rotate x-axis labels for better readability
plt.show()

我们也可以使用

np.where

方法。我们可以使用

plt.scatter

来实现这一点，如下面的代码：

import numpy as np
import pandas as pd
import matplotlib.pyplot as plt

# Sample DataFrame
data = {
    'timepoint': ['2021-01-01', '2021-02-01', '2021-11-01', '2021-12-01', '2022-01-01'],
    'value': [10, 20, 10, 50, 60],
    'point': [0, 0, 0, 1, 1]
}

df = pd.DataFrame(data)

# Convert 'timepoint' column to datetime type
df['timepoint'] = pd.to_datetime(df['timepoint'])

# Use np.where to conditionally set colors
col = np.where(df['point'] == 0, 'b', 'r')

# Create the scatter plot with conditional colors
plt.scatter(df['timepoint'], df['value'], c=col, s=50)

# Show the plot
plt.xlabel('Timepoint')
plt.ylabel('Value')
plt.title('Conditional Scatter Plot')
plt.xticks(rotation=45)  # Rotate x-axis labels for better readability
plt.grid(True)
plt.show()

这是创建线图的更新代码，但使用循环来区分颜色：

import numpy as np
import pandas as pd
import matplotlib.pyplot as plt

# Sample DataFrame
data = {
    'timepoint': ['2021-01-01', '2021-02-01', '2021-11-01', '2021-12-01', '2022-01-01'],
    'value': [10, 20, 10, 50, 60],
    'point': [0, 0, 0, 1, 1]
}

df = pd.DataFrame(data)

# Convert 'timepoint' column to datetime type
df['timepoint'] = pd.to_datetime(df['timepoint'])

# Use np.where to conditionally set colors
col = np.where(df['point'] == 0, 'b', 'r')

# Create the line graph with conditional colors
fig, ax = plt.subplots()

for color, group in df.groupby(col):
    ax.plot(group['timepoint'], group['value'], label=color)

# Add labels, legend, and show the plot
ax.set_xlabel('Timepoint')
ax.set_ylabel('Value')
plt.xticks(rotation=45)  # Rotate x-axis labels for better readability
plt.grid(True)
plt.legend()
plt.show()

你可以查看我的kaggle这里