数据可视化｜用箱线图进行对比分析

# 导入所需的库import numpy as npimport pandas as pdimport matplotlib as mplimport matplotlib.pyplot as plt# 正常显示中文标签mpl.rcParams["font.sans-serif"] = ["SimHei"]# 自动适应布局mpl.rcParams.update({"figure.autolayout": True})# 正常显示负号mpl.rcParams["axes.unicode_minus"] = False# 定义颜色，主色：蓝色，辅助色：灰色，互补色：橙色c = {"蓝色":"#00589F", "深蓝色":"#003867", "浅蓝色":"#5D9BCF",     "灰色":"#999999", "深灰色":"#666666", "浅灰色":"#CCCCCC",     "橙色":"#F68F00", "深橙色":"#A05D00", "浅橙色":"#FBC171"}

# 数据源路径filepath="./data/问卷调查结果3.xlsx"# 读取 Excel文件df = pd.read_excel(filepath, index_col="调查用户")# 定义画图用的数据labels = df.columnsdata = df.values

# 使用「面向对象」的方法画图，定义图片的大小fig, ax = plt.subplots(figsize=(8, 6))# 设置背景颜色fig.set_facecolor("w")ax.set_facecolor("w")# 设置箱体的属性boxprops = {"color":c["灰色"]}# 设置中位线参数medianprops = {"color":c["橙色"]}# 设置均值的属性meanprops = {"color":c["灰色"]}# 设置箱线图顶端和末端线条的属性capprops = {"color":c["灰色"]}# 设置须线的属性whiskerprops = {"color":c["灰色"]}# 绘制箱线图，whis 用来控制箱须包含数据的范围，这里设置得比较大，是为了不让图中出现异常点bplot = ax.boxplot(data, vert=False, patch_artist=True, labels=labels, showmeans=True, whis=5, meanline=True,                    boxprops=boxprops, medianprops=medianprops, meanprops=meanprops, capprops=capprops, whiskerprops=whiskerprops)# 设置标题ax.set_title("
用户满意度的分布情况
", loc="left", size=26, color=c["深灰色"])# 倒转 Y 轴，让第一个功能排在最上面ax.invert_yaxis()# 隐藏 Y 轴的刻度线ax.tick_params(axis="y", which="major", length=0)# 设置 X 轴范围ax.set_xlim(0.76, 1.02)# 隐藏 X 轴ax.xaxis.set_visible(False)# 标记箱线代表的含义ax.text(min(data[:,0]), 1, "%.1f%% "%(min(data[:,0])*100)+"
最小值 ", ha="right", fontsize=12, color=c["深灰色"], va="center")ax.text(np.mean(data[:,0]), 0.7, " %.1f%%"%(np.mean(data[:,0])*100)+"
 平均值", ha="center", fontsize=12, color=c["深灰色"], va="bottom")ax.text(np.median(data[:,0]), 1.3, " %.1f%%"%(np.median(data[:,0])*100)+"
中位数", ha="center", fontsize=12, color=c["橙色"], va="top")ax.text(max(data[:,0]), 1, " %.1f%%"%(max(data[:,0])*100)+"
 最大值", ha="left", fontsize=12, color=c["深灰色"], va="center")# 循环添加数据标签for i in range(1, len(labels)):    ax.text(min(data[:,i]), i+1, "%.1f%% "%(min(data[:,i])*100), ha="right", fontsize=12, color=c["深灰色"], va="center")    ax.text(np.mean(data[:,i]), i+0.7, " %.1f%%"%(np.mean(data[:,i])*100), ha="center", fontsize=12, color=c["深灰色"], va="bottom")    ax.text(np.median(data[:,i]), i+1.3, " %.1f%%"%(np.median(data[:,i])*100), ha="center", fontsize=12, color=c["橙色"], va="top")    ax.text(max(data[:,i]), i+1, " %.1f%%"%(max(data[:,i])*100), ha="left", fontsize=12, color=c["深灰色"], va="center")# 填充箱体的颜色colors = [c["浅橙色"], c["浅橙色"], c["浅灰色"], c["蓝色"], c["蓝色"]]for patch, color in zip(bplot["boxes"], colors):    patch.set_facecolor(color)# 隐藏边框ax.spines["top"].set_visible(False)ax.spines["right"].set_visible(False)ax.spines["bottom"].set_visible(False)ax.spines["left"].set_visible(False)# 设置坐标标签字体大小和颜色ax.tick_params(labelsize=16, colors=c["深灰色"])# 用百分比格式显示坐标轴def to_percent(temp, position=1):    return "%.0f"%(100 * temp) + "%"formatter = mpl.ticker.FuncFormatter(to_percent)ax.xaxis.set_major_formatter(formatter)plt.savefig("用户满意度的分布情况.jpeg")plt.show()