如何移除Matplotlib中多变量时间序列子图间的空白区域

要保留坐标轴的同时消除子图间的空白，只需对代码做以下3处修改：

• 创建子图时通过gridspec_kw设置垂直间距为0
• 隐藏上方子图的x轴刻度与标签，避免和下方子图重叠
• 调用tight_layout()优化整体布局，防止元素被裁剪

修改后的完整代码如下：

#@title Animate Define

# Display Parameter
start = 0
end = 1024

Xt = X_unseen[start:end]
yp = y_pred[start:end]
yt = y_true[start:end]

# Get mask of every class for prediction
bl_pred = yp == 0
p_pred = yp == 1
qrs_pred = yp == 2
t_pred = yp == 3

# get mask of every class for ground truth
bl_true = yt == 0
p_true = yt == 1
qrs_true = yt == 2
t_true = yt == 3

# 修改1：设置子图垂直间距为0
fig, (ax1, ax2) = plt.subplots(2, 1, figsize=(20, 8), gridspec_kw={'hspace': 0})

# Plotting for prediction
prev_class = None
start_idx = start
for i in range(start, end):
    current_class = None
    if bl_pred[i]:
        current_class = 'grey'
    elif p_pred[i]:
        current_class = 'orange'
    elif qrs_pred[i]:
        current_class = 'green'
    elif t_pred[i]:
        current_class = 'purple'

    if current_class != prev_class:
        if prev_class is not None:
            ax1.axvspan(start_idx, i, color=prev_class, alpha=0.5)
        start_idx = i
        prev_class = current_class
# Fill the last region
if prev_class is not None:
    ax1.axvspan(start_idx, end, color=prev_class, alpha=0.5)

# Plotting for ground truth
prev_class = None
start_idx = start
for i in range(start, end):
    current_class = None
    if bl_true[i]:
        current_class = 'grey'
    elif p_true[i]:
        current_class = 'orange'
    elif qrs_true[i]:
        current_class = 'green'
    elif t_true[i]:
        current_class = 'purple'

    if current_class != prev_class:
        if prev_class is not None:
            ax2.axvspan(start_idx, i, color=prev_class, alpha=0.5)
        start_idx = i
        prev_class = current_class
# Fill the last region
if prev_class is not None:
    ax2.axvspan(start_idx, end, color=prev_class, alpha=0.5)

# first row for prediction (X_unseen, y_pred)
ax1.plot(Xt, color='blue')
ax1.set_title('Prediction')
# 修改2：隐藏上方子图的x轴刻度和标签，避免重叠
ax1.set_xticks([])
ax1.set_xticklabels([])

# second row for ground truth (X_unseen, y_true)
ax2.plot(Xt, color='blue')
ax2.set_title('Ground Truth')

# 修改3：优化整体布局
plt.tight_layout()
# plot
plt.show()

内容的提问来源于stack exchange，提问作者Muhammad Ikhwan Perwira