IIabm/anova_visualization.ipynb

In [1]:

import pandas as pd
import matplotlib.pyplot as plt
import seaborn as sns

config = {"figure.dpi": 300,
          "font.family": 'serif',
          "font.serif": ['SimSun']}
df = pd.read_csv('analysis/20230818anova_visualization.csv', encoding='utf-8-sig')
df['sort_index'] = df['level'].map({'不倾向':0,
                                    '倾向':1,
                                    '低':0,
                                    '中':1,
                                    '高':2,
                                    '单供应商':0,
                                    '双供应商':1,
                                    '三供应商':2})
df.sort_values(['自变量', 'sort_index'], inplace=True)
df.drop(columns='sort_index', inplace=True)
df

Out [1]:

	自变量	level	系统恢复用时R1	产业-企业边累计扰乱次数R2	产业-企业边最大传导深度R3	产业-企业边断裂总数R4
15	新供应关系构成概率P7	低	2.240	2.672	1.143	0.7640
16	新供应关系构成概率P7	中	2.132	2.674	1.143	0.7859
17	新供应关系构成概率P7	高	2.179	2.649	1.124	0.7575
9	是否已有连接偏好P4	不倾向	2.177	2.668	1.141	0.7804
8	是否已有连接偏好P4	倾向	2.191	2.663	1.133	0.7579
4	是否规模偏好P2	不倾向	2.171	2.669	1.137	0.7726
3	是否规模偏好P2	倾向	2.196	2.661	1.137	0.7657
18	最大尝试时间步P8	低	1.726	2.646	1.123	0.7782
19	最大尝试时间步P8	中	2.186	2.682	1.144	0.7599
20	最大尝试时间步P8	高	2.640	2.667	1.143	0.7694
7	最大尝试次数P3	低	2.286	2.691	1.154	0.8254
6	最大尝试次数P3	中	2.124	2.652	1.127	0.7431
5	最大尝试次数P3	高	2.141	2.652	1.130	0.7390
2	采购策略P1	单供应商	2.261	2.519	1.121	0.7919
1	采购策略P1	双供应商	2.146	2.650	1.133	0.7615
0	采购策略P1	三供应商	2.144	2.826	1.156	0.7541
10	额外产能分布P5	均匀分布	2.316	2.681	1.158	0.8403
11	额外产能分布P5	正态分布	2.052	2.650	1.115	0.6980
14	额外产能分布参数P6	低	2.436	2.705	1.171	0.9121
13	额外产能分布参数P6	中	2.202	2.666	1.142	0.7655
12	额外产能分布参数P6	高	1.914	2.624	1.098	0.6299

In [2]:

x_name = '采购策略P1'
y_choose = [0, 1, 2]
y_prop = pd.DataFrame({'y_name': ['系统恢复用时R1', '产业-企业边累计扰乱次数R2', '产业-企业边最大传导深度R3', '产业-企业边断裂总数R4'],
                       'line_style': [(1, 0),(3, 1), (1,1), (3,2,1,2)],
                       'palette': sns.color_palette("deep")[0:4]})
df_x = df.loc[df['自变量'] == x_name, 'level':].set_index('level').stack(
).reset_index().rename(columns={'level': '水平', 'level_1': '响应变量', 0: '均值'})
df_x = df_x.loc[df_x['响应变量'].isin(y_prop.loc[y_choose]['y_name'])]
sns.set_theme(style="whitegrid", rc=config)
ax = sns.lineplot(data=df_x, x="水平", y="均值", hue="响应变量", style="响应变量",
                  markers=['o'],
                  dashes=y_prop.loc[y_choose]['line_style'].to_list(),
                  palette=y_prop.loc[y_choose]['palette'].to_list(),
                  legend='brief')
ax.set_title(x_name)
for item in df_x.groupby('响应变量'):
    for x, y, m in item[1][['水平', '均值', '均值']].values:
        ax.text(x, y+0.05, f'{m:.2f}')

C:\Users\ASUS\AppData\Local\Temp\ipykernel_42720\1808291987.py:10: UserWarning: 
The markers list has fewer values (1) than needed (3) and will cycle, which may produce an uninterpretable plot.
  ax = sns.lineplot(data=df_x, x="水平", y="均值", hue="响应变量", style="响应变量",

In [3]:

x_name = '最大尝试次数P3'
y_choose = [0, 1, 2, 3]
y_prop = pd.DataFrame({'y_name': ['系统恢复用时R1', '产业-企业边累计扰乱次数R2', '产业-企业边最大传导深度R3', '产业-企业边断裂总数R4'],
                       'line_style': [(1, 0),(3, 1), (1,1), (3,2,1,2)],
                       'palette': sns.color_palette("deep")[0:4]})
df_x = df.loc[df['自变量'] == x_name, 'level':].set_index('level').stack(
).reset_index().rename(columns={'level': '水平', 'level_1': '响应变量', 0: '均值'})
df_x = df_x.loc[df_x['响应变量'].isin(y_prop.loc[y_choose]['y_name'])]
sns.set_theme(style="whitegrid", rc=config)
ax = sns.lineplot(data=df_x, x="水平", y="均值", hue="响应变量", style="响应变量",
                  markers=['o'],
                  dashes=y_prop.loc[y_choose]['line_style'].to_list(),
                  palette=y_prop.loc[y_choose]['palette'].to_list(),
                  legend='brief')
ax.set_title(x_name)
for item in df_x.groupby('响应变量'):
    for x, y, m in item[1][['水平', '均值', '均值']].values:
        ax.text(x, y+0.05, f'{m:.2f}')

C:\Users\ASUS\AppData\Local\Temp\ipykernel_42720\1224603408.py:10: UserWarning: 
The markers list has fewer values (1) than needed (4) and will cycle, which may produce an uninterpretable plot.
  ax = sns.lineplot(data=df_x, x="水平", y="均值", hue="响应变量", style="响应变量",

In [4]:

x_name = '额外产能分布P5'
display_name = '额外产能分布类型P5'
y_choose = [0, 1, 2, 3]
y_prop = pd.DataFrame({'y_name': ['系统恢复用时R1', '产业-企业边累计扰乱次数R2', '产业-企业边最大传导深度R3', '产业-企业边断裂总数R4'],
                       'line_style': [(1, 0),(3, 1), (1,1), (3,2,1,2)],
                       'palette': sns.color_palette("deep")[0:4]})
df_x = df.loc[df['自变量'] == x_name, 'level':].set_index('level').stack(
).reset_index().rename(columns={'level': '水平', 'level_1': '响应变量', 0: '均值'})
df_x = df_x.loc[df_x['响应变量'].isin(y_prop.loc[y_choose]['y_name'])]
sns.set_theme(style="whitegrid", rc=config)
ax = sns.lineplot(data=df_x, x="水平", y="均值", hue="响应变量", style="响应变量",
                  markers=['o'],
                  dashes=y_prop.loc[y_choose]['line_style'].to_list(),
                  palette=y_prop.loc[y_choose]['palette'].to_list(),
                  legend='brief')
ax.set_title(display_name)
for item in df_x.groupby('响应变量'):
    for x, y, m in item[1][['水平', '均值', '均值']].values:
        ax.text(x, y+0.05, f'{m:.2f}')

C:\Users\ASUS\AppData\Local\Temp\ipykernel_42720\2665207915.py:10: UserWarning: 
The markers list has fewer values (1) than needed (4) and will cycle, which may produce an uninterpretable plot.
  ax = sns.lineplot(data=df_x, x="水平", y="均值", hue="响应变量", style="响应变量",

In [5]:

x_name = '额外产能分布参数P6'
display_name = '额外产能分布均值P6'
y_choose = [0, 1, 2, 3]
y_prop = pd.DataFrame({'y_name': ['系统恢复用时R1', '产业-企业边累计扰乱次数R2', '产业-企业边最大传导深度R3', '产业-企业边断裂总数R4'],
                       'line_style': [(1, 0),(3, 1), (1,1), (3,2,1,2)],
                       'palette': sns.color_palette("deep")[0:4]})
df_x = df.loc[df['自变量'] == x_name, 'level':].set_index('level').stack(
).reset_index().rename(columns={'level': '水平', 'level_1': '响应变量', 0: '均值'})
df_x = df_x.loc[df_x['响应变量'].isin(y_prop.loc[y_choose]['y_name'])]
sns.set_theme(style="whitegrid", rc=config)
ax = sns.lineplot(data=df_x, x="水平", y="均值", hue="响应变量", style="响应变量",
                  markers=['o'],
                  dashes=y_prop.loc[y_choose]['line_style'].to_list(),
                  palette=y_prop.loc[y_choose]['palette'].to_list(),
                  legend='brief')
ax.set_title(display_name)
for item in df_x.groupby('响应变量'):
    for x, y, m in item[1][['水平', '均值', '均值']].values:
        ax.text(x, y+0.05, f'{m:.2f}')

C:\Users\ASUS\AppData\Local\Temp\ipykernel_42720\759130329.py:10: UserWarning: 
The markers list has fewer values (1) than needed (4) and will cycle, which may produce an uninterpretable plot.
  ax = sns.lineplot(data=df_x, x="水平", y="均值", hue="响应变量", style="响应变量",

In [6]:

x_name = '新供应关系构成概率P7'
y_choose =[1]
y_prop = pd.DataFrame({'y_name': ['系统恢复用时R1', '产业-企业边累计扰乱次数R2', '产业-企业边最大传导深度R3', '产业-企业边断裂总数R4'],
                       'line_style': [(1, 0),(3, 1), (1,1), (3,2,1,2)],
                       'palette': sns.color_palette("deep")[0:4]})
df_x = df.loc[df['自变量'] == x_name, 'level':].set_index('level').stack(
).reset_index().rename(columns={'level': '水平', 'level_1': '响应变量', 0: '均值'})
df_x = df_x.loc[df_x['响应变量'].isin(y_prop.loc[y_choose]['y_name'])]
sns.set_theme(style="whitegrid", rc=config)
ax = sns.lineplot(data=df_x, x="水平", y="均值", hue="响应变量", style="响应变量",
                  markers=['o'],
                  dashes=y_prop.loc[y_choose]['line_style'].to_list(),
                  palette=y_prop.loc[y_choose]['palette'].to_list(),
                  legend='brief')
ax.set_title(x_name)
for item in df_x.groupby('响应变量'):
    for x, y, m in item[1][['水平', '均值', '均值']].values:
        ax.text(x, y+0.0005, f'{m:.2f}')

In [7]:

x_name = '最大尝试时间步P8'
y_choose=[0,1,2]
y_prop = pd.DataFrame({'y_name': ['系统恢复用时R1', '产业-企业边累计扰乱次数R2', '产业-企业边最大传导深度R3', '产业-企业边断裂总数R4'],
                       'line_style': [(1, 0),(3, 1), (1,1), (3,2,1,2)],
                       'palette': sns.color_palette("deep")[0:4]})
df_x = df.loc[df['自变量'] == x_name, 'level':].set_index('level').stack(
).reset_index().rename(columns={'level': '水平', 'level_1': '响应变量', 0: '均值'})
df_x = df_x.loc[df_x['响应变量'].isin(y_prop.loc[y_choose]['y_name'])]
sns.set_theme(style="whitegrid", rc=config)
ax = sns.lineplot(data=df_x, x="水平", y="均值", hue="响应变量", style="响应变量",
                  markers=['o'],
                  dashes=y_prop.loc[y_choose]['line_style'].to_list(),
                  palette=y_prop.loc[y_choose]['palette'].to_list(),
                  legend='brief')
ax.set_title(x_name)
for item in df_x.groupby('响应变量'):
    for x, y, m in item[1][['水平', '均值', '均值']].values:
        ax.text(x, y+0.05, f'{m:.2f}')

C:\Users\ASUS\AppData\Local\Temp\ipykernel_42720\1838672856.py:10: UserWarning: 
The markers list has fewer values (1) than needed (3) and will cycle, which may produce an uninterpretable plot.
  ax = sns.lineplot(data=df_x, x="水平", y="均值", hue="响应变量", style="响应变量",

In [8]:

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

# 风格参考 analysis_firm_risk_component.ipynb
matplotlib.rcParams['font.sans-serif'] = ['SimSun']
matplotlib.rcParams['axes.unicode_minus'] = False

x_names = ['采购策略P1', '最大尝试时间步P8']
fig, axes = plt.subplots(1, 2, figsize=(6, 3), dpi=300, sharey=True)

for idx, x_name in enumerate(x_names):
    ax = axes[idx]

    df_x = df.loc[df['自变量'] == x_name, 'level':].set_index('level').stack(
    ).reset_index().rename(columns={'level': '水平', 'level_1': '响应变量', 0: '均值'})

    df_plot = df_x.loc[df_x['响应变量'] == '系统恢复用时R1'].copy()
    df_plot['响应变量'] = '产业链供应链恢复时间'

    x_pos = np.arange(len(df_plot))
    y_vals = df_plot['均值'].values
    ax.plot(x_pos, y_vals, marker='o', linewidth=2, alpha=0.85, color='C0')

    # y_offset = (y_vals.max() - y_vals.min()) * 0.02 if len(y_vals) > 0 else 0.05
    # if y_offset == 0:
    #     y_offset = 0.05
    # for x, y in zip(x_pos, y_vals):
    #     ax.text(x, y + y_offset, f'{y:.2f}', ha='center', va='bottom', fontsize=8)

    ax.set_xlabel(x_name, fontsize=11)
    ax.set_xticks(x_pos)
    ax.set_xticklabels(df_plot['水平'].astype(str).values)

    if idx == 0:
        ax.set_ylabel('产业链供应链恢复时间', fontsize=11)

    ax.grid(True, alpha=0.3)

    for spine in ax.spines.values():
        spine.set_edgecolor('black')
        spine.set_linewidth(1.0)

plt.tight_layout()
plt.show()

In [10]:

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

# 风格参考 analysis_firm_risk_component.ipynb
matplotlib.rcParams['font.sans-serif'] = ['SimSun']
matplotlib.rcParams['axes.unicode_minus'] = False

x_names = ['最大尝试次数P3', '额外产能分布P5', '额外产能分布参数P6']
x_labels = {'额外产能分布P5': '额外产能分布类型P5', '额外产能分布参数P6': '额外产能分布均值P6'}
fig, axes = plt.subplots(1, 3, figsize=(7, 3), dpi=300, sharey=True)

for idx, x_name in enumerate(x_names):
    ax = axes[idx]

    df_x = df.loc[df['自变量'] == x_name, 'level':].set_index('level').stack(
    ).reset_index().rename(columns={'level': '水平', 'level_1': '响应变量', 0: '均值'})

    df_plot = df_x.loc[df_x['响应变量'] == '系统恢复用时R1'].copy()
    df_plot['响应变量'] = '产业链供应链恢复时间'

    x_pos = np.arange(len(df_plot))
    y_vals = df_plot['均值'].values
    ax.plot(x_pos, y_vals, marker='o', linewidth=2, alpha=0.85, color='C0')

    # y_offset = (y_vals.max() - y_vals.min()) * 0.02 if len(y_vals) > 0 else 0.05
    # if y_offset == 0:
    #     y_offset = 0.05
    # for x, y in zip(x_pos, y_vals):
    #     ax.text(x, y + y_offset, f'{y:.2f}', ha='center', va='bottom', fontsize=8)

    ax.set_xlabel(x_labels.get(x_name, x_name), fontsize=11)
    ax.set_xticks(x_pos)
    ax.set_xticklabels(df_plot['水平'].astype(str).values)

    if idx == 0:
        ax.set_ylabel('产业链供应链恢复时间', fontsize=11)

    ax.grid(True, alpha=0.3)

    for spine in ax.spines.values():
        spine.set_edgecolor('black')
        spine.set_linewidth(1.0)

plt.tight_layout()
plt.show()

In [ ]: