Merge branch 'sebl' into 'main'

Calculations.py goes through all speed-limit zones and calculates how many...

See merge request dbis/lecture-groups/database-systems/2023hs/group-1!6

analysis/calculations.py

@@ -0,0 +1,68 @@
+import logging
+from shapely.geometry import Point, LineString
+from shapely import wkb
+from db_connector import RemoteDB
+import pandas as pd
+
+speedLimits = ["T0", "T20", "T30", "T50","T60", "T80", "T100"]
+
+
+def is_point_near_multilinestring(point, multilinestring, threshold_distance):
+    point_geometry = Point(point)
+    return point_geometry.distance(multilinestring) < threshold_distance
+
+
+def get_data(db):
+    get_speeds_sql = """
+    SELECT wkb_geometry,
+       temporegime_technical
+    FROM signaled_speeds;
+    """
+
+    result = db.execute_query(get_speeds_sql)
+    sig_speed_df = pd.DataFrame(result)
+    sig_speed_df.rename(columns={'wkb_geometry': 'geometry'}, inplace=True)
+    sig_speed_df['geometry'] = sig_speed_df['geometry'].apply(lambda x: wkb.loads(x, hex=True))
+
+    get_accidents = """
+    SELECT geometry
+    FROM accidents;
+    """
+    result = db.execute_query(get_accidents)
+    accident_df = pd.DataFrame(result)
+    accident_df['geometry'] = accident_df['geometry'].apply(lambda x: wkb.loads(x, hex=True))
+
+    process_data(sig_speed_df, accident_df)
+
+
+def process_data(sig_speed_df, accident_df):
+    result_df = pd.DataFrame(columns= ['TempoLim', 'Accidents_total'])
+    for speed in speedLimits:
+        print("Checking for zone: " + speed)
+        filtered_df = sig_speed_df[sig_speed_df["temporegime_technical"].str.contains(speed, case=False, na=False)]
+        current_result = count_points_near_multilinestrings(accident_df, filtered_df, 0.000005)
+        result_df.loc[len(result_df)] = {'TempoLim': speed, 'Accidents_total': current_result}
+    print("FINAL RESULT")
+    print(result_df)
+
+
+def count_points_near_multilinestrings(points_df, multilinestrings_df, threshold_distance):
+    result_counts = []
+
+    for idx, multilinestring_row in multilinestrings_df.iterrows():
+        multilinestring = multilinestring_row['geometry']
+        count_near = sum(points_df['geometry'].apply(
+            lambda point: is_point_near_multilinestring(point, multilinestring, threshold_distance)))
+        result_counts.append({'temporegime_technical': multilinestring_row['temporegime_technical'], 'CountNear': count_near})
+    result_df = pd.DataFrame(result_counts)
+    return result_df['CountNear'].sum()
+
+
+if __name__ == "__main__":
+    remote_db = RemoteDB()
+    try:
+        get_data(remote_db)
+    except Exception as e:
+        print(f"Exception {e} in calculations.py")
+    finally:
+        remote_db.close()

analysis/config_example.py

@@ -1,5 +1,5 @@
 # config.py, adjust as needed
-# TODO RENAME THIS FILE TO "config.py"
+# TODO COPY and then RENAME TO "config.py"
 SSH_HOST = 'slenzlinger.dev'
 SSH_USERNAME = 'sebl'  #TODO: Enter own username
 SSH_PASSWORD = 'your_ssh_password' # TODO: to not push to git

analysis/map.py

@@ -300,4 +300,4 @@ if __name__ == "__main__":
 
     ## Save Maps ============================================================================================
     save_map_as_html(toggle_map, "heat_map_toggle")
-    save_map_as_html(time_map, "heat_map_time")
+    save_map_as_html(time_map, "html/heat_map_time")

analysis/plots.py

@@ -10,8 +10,6 @@ import numpy as np
 logging.getLogger("matplotlib").setLevel(logging.WARNING)
 
 
-
-
 # Summary charts ======================================================================================================
 def plt_acc_by_year(db):
     acc_year_sql = """
@@ -73,6 +71,7 @@ def plt_acc_by_day_year(db):
         animation_frame='year',
         labels={'weekday': 'Weekday', 'count': 'Number of Accidents'},
         category_orders={'weekday': ['Saturday', 'Sunday', 'Monday', 'Tuesday', 'Wednesday', 'Thursday', 'Friday']},
+        orientation='h'
     )
     fig.update_yaxes(range=[0, 1000])
     # Customize the layout to include a slider
@@ -137,9 +136,9 @@ def plt_acc_by_daytime(db):
     result = db.execute_query(acc_weekday_sql)
     result_df = pd.DataFrame(result)
 
-    fig = px.bar(result_df, y='hour', x='count', orientation='h')
+    fig = px.bar(result_df, y='hour', x='count', orientation='h', title='Accidents by day')
-    fig.write_image("fig/acc_by_day.png")
+    fig.write_image("fig/acc_by_daytime.png")
-    fig.write_html("html/acc_by_day.html")
+    fig.write_html("html/acc_by_daytime.html")
 
 # Time Series charts ==================================================================================================
 def acc_by_type(db):
@@ -260,8 +259,6 @@ def severity_by_month(db):
     #fig.show()
 
 
-
-
 # Utilities ===========================================================================================================
 def save_as_barplot(df, xname, yname, orientation, file_name):
     pass