Upload 4 files

ev_merged_analysis_readme.md

@@ -0,0 +1,29 @@
+# Full EV Charging Station Infrastructure EDA
+## Overview about the dataset
+- A merged dataset contains information about most ev charging stations in Finland using 2 different sources: from Fintraffic and from Google.
+- For more information regarding the details about the pre-merged sets, the merging process, and overview about the current set, check out the other README file in the same branch.
+## Visualization
+- Contains an interactable map and a simple histogram.
+- Each dot on the map represents a station, the size of the dot correlates to the number of connectors that the station has. Hover on the dot gives information regarding its name, operator name, total connectors, municipality, total charging posts (evses), and max power capacity.
+- Histogram shows top 20 cities with most stations.
+- To interact with the map, run the code on your environment.
+- The station is categorised into 3 tiers: slow, fast, and ultra fast, represented by 3 different colors. This is done through using the highest capacity connector's powert as the divider, with breakpoints given from official EU sources:
+    + <50kw: slow
+    + 50-149kw: fast
+    + from 150kw: ultra fast
+- Intepretation:
+    + Map shows that the majority of stations only contain slow connectors, meaning that we do not necessarily have to waste budget on building the fastest charging station.
+    + Most slow stations are clustered in big cities such as Helsinki, Espoo, Tampere.
+    + Ultra fast stations are found scattering across all of Finland, especially in less urban areas. This is explained by drivers' stronger needs for faster charging speed in secluded areas, where not many stations are present.
+    + Throughout the highways and especially on intersection of them, fast and ultra fast stations are usually present.
+    + Histogram supports the interpretations, with big cities dominate and a majority of slow charging stations.
+    + Interestingly, most cities contain more ultra fast charging stations rather than just fast (also supported by the map). This could be interpreted as fast charging stations are more niche and usually does not fit the situations that people would use the charging stations.
+## Compatibility with other dataset
+- Any dataset with clear coordinates that can be represented on the map is a possible candidate and should be analysed with this for a clearer understanding of what kind of places do current operators place their stations.
+- Specifically, the point of interest dataset is highly compatible with this dataset and should also be represented on the map to analyse correlation.
+## Future steps
+- Final cleaning:
+    + Removing redundant/unnecessary columns.
+    + Handling null values (columns such as userRating while interesting has 70% null entries).
+- Combine with the POI dataset (and possibly other dataset with clear coordinates) for further analysis.
+

ev_merged_eda.py

@@ -0,0 +1,79 @@
+import pandas as pd
+import numpy as np
+import plotly.express as px
+import geopandas as gpd
+from shapely.geometry import Point
+
+df = pd.read_csv('ev_stations_merged.csv')
+print(df.head())
+print(df.columns)
+print(df.isna().sum())
+print(df.describe())
+print(df['place_key'].duplicated().sum())
+
+municipalities = gpd.read_file('kuntarajat.2023.json')
+geometry = [Point(xy) for xy in zip(df['lng'], df['lat'])]
+gdf = gpd.GeoDataFrame(df, geometry=geometry, crs='EPSG:4326')
+municipalities = municipalities.to_crs('EPSG:4326')
+joined_df = gpd.sjoin(gdf, municipalities, how='left')
+print(joined_df.columns)
+df['municipality'] = joined_df['NAMEFIN']
+print(df['municipality'].isna().sum())
+df = df[df['municipality'].notna()].copy()
+print(df.isna().sum())
+
+
+def charger_speed_categorised(power):
+    if pd.isna(power):
+        return np.nan
+    elif power < 50:
+        return 'slow'
+    elif 50 <= power < 150:
+        return 'fast'
+    else:
+        return 'ultra_fast'
+
+df['charger_speed_categorised'] = df['max_power_kw'].apply(charger_speed_categorised)
+
+fig_map = px.scatter_map(
+    df, 
+    lat='lat',
+    lon='lng',
+    color='charger_speed_categorised',
+    size='total_connectors',
+    size_max=30,
+    zoom=5,
+    center={'lat': 64.23, 'lon': 27.68},
+    hover_name='name',
+    hover_data={
+        'municipality': True,
+        'operator_name': True,
+        'max_power_kw': True,
+        'total_evses': True,
+        'lat': False,
+        'lng': False
+    },
+    map_style='open-street-map',
+    title='EV Charging Stations in Finland by Charging Speed'
+)
+fig_map.show()
+
+municipality_summary = df.groupby('municipality').agg(
+    total_stations=('place_key','count'),
+    total_connectors=('total_connectors','sum'),
+    mean_power=('max_power_kw','mean')
+).sort_values('total_stations', ascending=False)
+
+top_20_municipality = municipality_summary.index[:20].tolist()
+df_top_20 = df[df['municipality'].isin(top_20_municipality)]
+
+
+fig_hist_station = px.histogram(
+    df_top_20, x='municipality', 
+    color='charger_speed_categorised',
+    category_orders={'municipality': top_20_municipality}
+)
+fig_hist_station.show()
+
+print(df.describe())
+