考虑两个数据帧 df1 和 df2,它们具有相同的索引,但可能具有不同的列,其中每个数据帧行代表一个类别,每列代表一个指标。目标是创建聚类堆叠条形图,其中每个类别的条形图分组在一起,每个数据帧的条形图彼此堆叠。
<code class="python">import pandas as pd import matplotlib.pyplot as plt import matplotlib.cm as cm def plot_clustered_stacked(df_list, labels=None, title="Clustered Stacked Bar Plot"): n_dataframes = len(df_list) n_columns = len(df_list[0].columns) n_index = len(df_list[0].index) fig, ax = plt.subplots() # Iterate through each dataframe for i, df in enumerate(df_list): # Plot the bars for the current dataframe df.plot(kind="bar", ax=ax, linewidth=0, stacked=True, legend=False, grid=False) # Adjust the position and width of the bars for df, j in zip(df_list, range(n_dataframes)): for n, rect in enumerate(ax.patches): if rect.get_y() == 0: # Stacked bar for dataframe df rect.set_x(rect.get_x() + j / float(n_dataframes)) rect.set_width(1 / float(n_dataframes)) # Set the x-axis labels and ticks ax.set_xticks(np.arange(0, n_index) + 0.5) ax.set_xticklabels(df.index) # Add a legend for the dataframes plt.legend([df.stack(level=0).index[0] for df in df_list], labels) # Set the plot title ax.set_title(title) # Create example dataframes df1 = pd.DataFrame(np.random.rand(4, 3), index=["A", "B", "C", "D"], columns=["x", "y", "z"]) df2 = pd.DataFrame(np.random.rand(4, 3), index=["A", "B", "C", "D"], columns=["x", "y", "z"]) # Plot the clustered stacked bar plot plot_clustered_stacked([df1, df2], labels=["df1", "df2"])</code>
<code class="python">import seaborn as sns # Concatenate the dataframes into a single dataframe with a wide format df = pd.concat([df1.reset_index().melt(id_vars=["index"]), df2.reset_index().melt(id_vars=["index"])]) # Plot the clustered stacked bar plot g = sns.FacetGrid(data=df, col="variable", hue="index") g.map_dataframe(sns.barplot, order=df["index"].unique())</code>
以上是如何使用 Pandas 和 Matplotlib 或 Seaborn 创建聚类堆积条形图?的详细内容。更多信息请关注PHP中文网其他相关文章!