Added demo_results.csv file; changed PCI to be the abs value of the d…

…iff; updated plots

results/demo_results.csv

@@ -0,0 +1,24 @@
+Train Start,Train End,Pred Start,Pred End,Test F1,Pred F1
+2018-01-01,2022-01-01,2022-01-02,2022-04-01,0.615953947368421,0.4787234042553192
+2018-02-01,2022-02-01,2022-02-02,2022-05-01,0.6178199909543193,0.5253623188405797
+2018-03-01,2022-03-01,2022-03-02,2022-06-01,0.5974754558204769,0.5703422053231939
+2018-04-01,2022-04-01,2022-04-02,2022-07-01,0.6084257206208425,0.5390625
+2018-05-01,2022-05-01,2022-05-02,2022-08-01,0.6137767220902612,0.5656192236598891
+2018-06-01,2022-06-01,2022-06-02,2022-09-01,0.5934065934065934,0.5559655596555966
+2018-07-01,2022-07-01,2022-07-02,2022-10-01,0.5897542690545606,0.5067064083457526
+2018-08-01,2022-08-01,2022-08-02,2022-11-01,0.5713083718973496,0.5566714490674318
+2018-09-01,2022-09-01,2022-09-02,2022-12-01,0.6029055690072639,0.5910290237467019
+2018-10-01,2022-10-01,2022-10-02,2023-01-01,0.5700301204819277,0.5462962962962963
+2018-11-01,2022-11-01,2022-11-02,2023-02-01,0.590539894887721,0.5226130653266332
+2018-12-01,2022-12-01,2022-12-02,2023-03-01,0.5750337990085624,0.4334705075445816
+2019-01-01,2023-01-01,2023-01-02,2023-04-01,0.591604938271605,0.4949832775919732
+2019-02-01,2023-02-01,2023-02-02,2023-05-01,0.5755248335893497,0.4581939799331103
+2019-03-01,2023-03-01,2023-03-02,2023-06-01,0.5484913793103449,0.5085995085995086
+2019-04-01,2023-04-01,2023-04-02,2023-07-01,0.5816692268305171,0.5619596541786743
+2019-05-01,2023-05-01,2023-05-02,2023-08-01,0.5583153347732182,0.5398773006134969
+2019-06-01,2023-06-01,2023-06-02,2023-09-01,0.5728531855955679,0.513595166163142
+2019-07-01,2023-07-01,2023-07-02,2023-10-01,0.5687361419068736,0.5167464114832536
+2019-08-01,2023-08-01,2023-08-02,2023-11-01,0.5659369994660972,0.5530085959885387
+2019-09-01,2023-09-01,2023-09-02,2023-12-01,0.5542168674698795,0.5
+2019-10-01,2023-10-01,2023-10-02,2024-01-01,0.5565978736710444,0.5073313782991202
+2019-11-01,2023-11-01,2023-11-02,2024-02-01,0.5650557620817844,0.4870848708487085

src/analysis_util.py

@@ -86,12 +86,12 @@ def plot_results(df, event_dates, output_dir='../results/figures'):
 
     # Plot the results
     plt.figure(figsize=(12, 6), facecolor='white')
-    sns.set_style("white")  # 使用白色背景，不包含网格线
+    sns.set_style("white")  
 
     ax = sns.lineplot(data=plot_df, x='Pred End', y='Test F1', label='Test F1', color='navy')
     sns.lineplot(data=plot_df, x='Pred End', y='Pred F1', label='Pred F1', color='crimson')
 
-    # 移除网格线
+    # Remove grids
     ax.grid(False)
 
     # Add vertical lines for important events and event descriptions
@@ -105,30 +105,48 @@ def plot_results(df, event_dates, output_dir='../results/figures'):
         ax.text(event_date, event_position, event_desc, rotation=0, ha='left', va='center', fontsize=8)
 
     plt.title('Test and Prediction F1 Over Time', fontsize=16)
-    plt.xlabel('Prediction End Date', fontsize=14)
     plt.ylabel('F1 Score', fontsize=14)
-    plt.xticks(rotation=45)
-    ax.xaxis.set_major_formatter(mdates.DateFormatter('%Y-%m-%d'))
-
-    # Set x-axis ticks to every two months
+
+    # Set x-axis ticks to every two months and rotate labels
     ax.xaxis.set_major_locator(mdates.MonthLocator(interval=2))
     ax.xaxis.set_major_formatter(mdates.DateFormatter('%Y-%m-%d'))
+    plt.xticks(rotation=45, ha='right', rotation_mode='anchor')
 
     # Adjust the x-axis limits to show the desired time range
     ax.set_xlim([datetime(2022, 2, 1), datetime(2024, 4, 30)])
 
+    # Remove x-axis label
+    ax.set_xlabel('')
+
+    # Configure tick lines to protrude downward
+    ax.tick_params(
+        axis='x',
+        which='major',
+        direction='out',   # Tick direction outward (downward)
+        length=5,         # Length of major ticks
+        width=1,           # Width of tick lines
+        colors='black',    # Color of tick lines
+        pad=2,             # Distance between tick labels and tick lines
+        bottom=True,       # Show ticks at the bottom
+        top=False          # Don't show ticks at the top
+    )
+
+    # Ensure the bottom spine is visible and styled
+    ax.spines['bottom'].set_visible(True)
+    ax.spines['bottom'].set_linewidth(1)
+    ax.spines['bottom'].set_color('black')
+
     plt.tight_layout()  # Adjust the plot to ensure everything fits without overlapping
     plt.savefig(output_file_1, dpi=300, bbox_inches='tight', facecolor='white')  # Set facecolor to 'white'
     plt.show()
 
     # Second plot to show the difference between the test and prediction F1
     plt.figure(figsize=(12, 6), facecolor='white')
-    sns.set_style("white")  # 使用白色背景，不包含网格线
+    sns.set_style("white")  
 
     plot_df['F1 Diff'] = plot_df['Test F1'] - plot_df['Pred F1']
     ax = sns.lineplot(data=plot_df, x='Pred End', y='F1 Diff')
 
-    # 移除网格线
     ax.grid(False)
 
     # Add vertical lines for important events and event descriptions
@@ -143,20 +161,40 @@ def plot_results(df, event_dates, output_dir='../results/figures'):
 
     plt.ylabel('Policy Change Index for North Korea', fontsize=14)
     plt.axhline(0, color='red', linestyle='--')  # Add a horizontal line at 0
-    plt.xticks(rotation=45)
-
-    # Set x-axis ticks to every two months
+
+    # Set x-axis ticks to every two months and rotate labels
     ax.xaxis.set_major_locator(mdates.MonthLocator(interval=2))
     ax.xaxis.set_major_formatter(mdates.DateFormatter('%Y-%m-%d'))
+    plt.xticks(rotation=45, ha='right', rotation_mode='anchor')
 
     # Adjust the x-axis limits to show the desired time range
     ax.set_xlim([datetime(2022, 2, 1), datetime(2024, 4, 30)])
 
+    # Remove x-axis label
+    ax.set_xlabel('')
+
+    # Configure tick lines to protrude downward
+    ax.tick_params(
+        axis='x',
+        which='major',
+        direction='out',   # Tick direction outward (downward)
+        length=5,         # Length of major ticks
+        width=1,           # Width of tick lines
+        colors='black',    # Color of tick lines
+        pad=2,             # Distance between tick labels and tick lines
+        bottom=True,       # Show ticks at the bottom
+        top=False          # Don't show ticks at the top
+    )
+
+    # Ensure the bottom spine is visible and styled
+    ax.spines['bottom'].set_visible(True)
+    ax.spines['bottom'].set_linewidth(1)
+    ax.spines['bottom'].set_color('black')
+
     plt.tight_layout()
     plt.savefig(output_file_2, dpi=300, bbox_inches='tight', facecolor='white')  # Set facecolor to 'white'
     plt.show()
 
-
 ########################################################### 
 ## 2. LLM analysis
 

src/model.py

@@ -265,7 +265,7 @@ def main(args):
                          columns=['Train Start', 'Train End', 'Pred Start', 'Pred End', 'Test F1', 'Pred F1'])
 
     all_res = pd.concat([all_res, res], ignore_index=True)
-    all_res['PCI'] = (all_res['Test F1'] - all_res['Pred F1'])
+    all_res['PCI'] = abs((all_res['Test F1'] - all_res['Pred F1']))
     all_res.to_csv(all_res_file, index=False)
 
     # Print the full training history