需求： 使用Python的matplotlib绘制数据分布、数据箱型图、数据散点图
参考： https://blog.csdn.net/weixin_39559994/article/details/128197965?fromshare=blogdetail\&sharetype=blogdetail\&sharerId=128197965\&sharerefer=PC\&sharesource=captain_keating\&sharefrom=from_link
输入数据格式： df为dataframe格式，样例如下
输出效果：
代码/作图函数：
python
def plot_raincloud():
fig = plt.figure(figsize = (15,15),dpi = 300) # 创建Figure对象
plt.rcParams['font.sans-serif'] = 'SimHei' # 设置字体为SimHei显示中文在这里插入代码片，如果你的坐标轴有中文名，须设置
plt.rcParams['font.family'] = 'Arial' # 设置字体样式
plt.rcParams['font.size'] = '14' # 设置字体大小
plt.rcParams['xtick.direction'] = 'out' #将x周的刻度线方向设置向内
plt.rcParams['ytick.direction'] = 'out' #将y轴的刻度方向设置向内
colors = ['tomato', 'darksalmon', 'deepskyblue', 'mediumseagreen', 'orange','#bce27f','#aab8d8']
# 增加rain部分的随即抖动
data_x = [df[col].dropna().values for col in df.columns]
for i in range(len(data_x)):
idxs = np.arange(len(data_x[i]))
out = data_x[i].astype(float)
out.flat[idxs] += np.random.uniform(low=-0.1, high=0.1, size=len(idxs))
data_x[i] = out
# 计算统计信息
statistics = {
'Feature': [],'Mean': [],'Variance': [],'StdDev': [],'Min': [],
'Max': [],'25th Percentile': [],'50th Percentile (Median)': [],
'75th Percentile': [],'SEM': []
}
for idx, col in enumerate(new_column_names):
statistics['Feature'].append(col)
statistics['Mean'].append(np.mean(data_x[idx]))
statistics['Variance'].append(np.var(data_x[idx]))
statistics['StdDev'].append(np.std(data_x[idx]))
statistics['Min'].append(np.min(data_x[idx]))
statistics['Max'].append(np.max(data_x[idx]))
statistics['25th Percentile'].append(np.percentile(data_x[idx], 25))
statistics['50th Percentile (Median)'].append(np.percentile(data_x[idx], 50))
statistics['75th Percentile'].append(np.percentile(data_x[idx], 75))
# 计算SEM (标准误差)
n = len(data_x[idx]) # 样本数量
sem = np.std(data_x[idx]) / np.sqrt(n) # 计算SEM
statistics['SEM'].append(sem)
statistics_df = pd.DataFrame(statistics)
# 创建图片
ax = fig.add_axes([0.1, 0.61, 0.42, 0.3])
# ------------------------------
# 绘制 violin 图（雨图）
vp = ax.violinplot(data_x, points=500, showmeans=False, widths=1.1,# 控制小提琴图的宽度
showextrema=False, showmedians=False, vert=True)
for idx, b in enumerate(vp['bodies']):
b.set_color(colors[idx])
b.get_paths()[0].vertices[:, 0] = np.clip(b.get_paths()[0].vertices[:, 0], idx+1, idx+2)
# ------------------------------
# 绘制散点图（滴点）
for idx, y_vals in enumerate(data_x):
x = np.full(len(y_vals), idx + 0.8)
x += np.random.uniform(low=-0.15, high=0.15, size=len(x))# low和high调整散点图抖动范围
ax.scatter(x, y_vals, s=2, c=colors[idx], alpha=0.5)
# ------------------------------
# 绘制箱型图（箱图）
box_positions = np.arange(0.8, len(new_column_names)+0.8) # 控制箱型图的左右偏移
bp = ax.boxplot(data_x, vert=True, patch_artist=True, positions=box_positions,
widths=0.2, showfliers=False,# 控制箱型图宽度
boxprops=dict(facecolor='none', color='black'),
medianprops=dict(color='black'),
whiskerprops=dict(color='black'),
capprops=dict(color='black'))
# ---------------------------
ax.invert_yaxis()
ylim.reverse()
ax.set_ylim([-20,1100])
ax.set_xticks(np.arange(1, len(new_column_names)+1))
ax.set_xticklabels(new_column_names, rotation=60,fontdict={"family": "SimSun", "size": 20})
ax.set_ylabel('y轴标题',fontdict={"family": "SimSun", "size": 20})
ax.grid(True, linestyle='--', alpha=0.5)
return statistics_df
# 调用函数，返回统计指标dataframe
df1 = plot_raincloud() 备注：
new_column_names这个变量里装的是列名的列表，封装函数时忘了加进去了，大家使用的时候自己定义下即可，添加y坐标用的