According the article How Many People Die From Car Accidents Each Year, about 1.3 million death are caused by car accidents annually over the world. Moreover, it is stated that in the United States, 32 people die because of drunk driving car crashes per day. In my home country, public are so used to these kind of events that even many political figures were one of the perpetrators.
I have been wondering that why do car accidents still happen so regularly even if there are so many related laws. What kind of policy seems to address the problem the most? It seems like these answer could be found by compairing the accidents and the policy made in different region, and I hope to find a combination of policies that could lowest the car accident possibility.
import pandas as pd
df_traffic_death = pd.read_csv('traffic_death.csv')
traffic_death_feature = {'Entity': "Country", "Year": "years from 1990 to 2019",
"Deaths": "traffic deaths number"}
df_traffic_death.head(5)
| Entity | Year | Deaths | |
|---|---|---|---|
| 0 | Afghanistan | 1990 | 4154 |
| 1 | Afghanistan | 1991 | 4472 |
| 2 | Afghanistan | 1992 | 5106 |
| 3 | Afghanistan | 1993 | 5681 |
| 4 | Afghanistan | 1994 | 6001 |
df_death_type = pd.read_csv('death_type.csv')
death_type_feature = {'Country': 'Country',
"Drivers/passengers of 4-wheeled vehicles": "% of road traffic deaths by type",
"Drivers/passengers of motorized 2- or 3-wheelers": "% of road traffic deaths by type",
"Cyclists": "% of road traffic deaths by type",
"Pedestrians": "% of road traffic deaths by type",
"Other/unspecified roadusers": "% of road traffic deaths by type"}
df_death_type.head(5)
| Country | Drivers/passengers of 4-wheeled vehicles | Drivers/passengers of motorized 2- or 3-wheelers | Cyclists | Pedestrians | Other/unspecified road users | |
|---|---|---|---|---|---|---|
| 0 | Albania | 39.4 | 11.9 | 7.8 | 38.7 | 2.2 |
| 1 | Andorra | NaN | 50.0 | NaN | 50.0 | NaN |
| 2 | Angola | 59.5 | NaN | NaN | 40.5 | 0.0 |
| 3 | Antigua and Barbuda | 62.5 | 0.0 | 12.5 | 25.0 | 0.0 |
| 4 | Argentina | 47.2 | 22.2 | 2.4 | 8.2 | 20.0 |
df_drunk_rules = pd.read_csv('drunk_rule.csv')
drunk_rules_feature = {"Country": "Country",
"Definition of drink-driving by BAC": "whether or not the conutry have clear BAC to identify drink-driving",
"Existence of a national drink-drrving law": "whether or not having law about rink-driving",
"Attribution of road traffic deaths to alcohol": "% of traffic deaths caused by alcohol"}
df_drunk_rules.head(5)
| Country | Definition of drink-driving by BAC | Existence of a national drink-driving law | Attribution of road traffic deaths to alcohol (%) | |
|---|---|---|---|---|
| 0 | Afghanistan | No | Yes | – |
| 1 | Albania | Yes | Yes | 5.2 |
| 2 | Angola | Yes | Yes | – |
| 3 | Antigua and Barbuda | No | Yes | 0.9 |
| 4 | Argentina | Yes | Yes | 17 |
df_bac_limit = pd.read_csv('bac_limit.csv')
bac_limit_feature = {"Country": "Country",
"BAC limit for general population": "the blood alcohol concentration standard for most people",
"BAC limit for novice drivers": "the blood alcohol concentration standard for young or new drivers"}
df_bac_limit.head(5)
| Country | BAC limit for general population | BAC limit for novice drivers | |
|---|---|---|---|
| 0 | Afghanistan | - | - |
| 1 | Albania | <=0.05 g/dl | <=0.05 g/dl |
| 2 | Angola | <= 0.06 g/dl | <= 0.06 g/dl |
| 3 | Antigua and Barbuda | - | - |
| 4 | Argentina | <=0.05 g/dl | <=0.05 g/dl |
df_speed_limit = pd.read_csv('speed_limit.csv')
speed_limit_feature = {"Country": "Country",
"Maximum speed limits (Urban)": "speed limits in city side",
"Maximum speed limits (Rural)": "speed limits in country side"}
df_speed_limit.head(5)
| Country | Maximum speed limits (Urban) | Maximum speed limits (Rural) | |
|---|---|---|---|
| 0 | Afghanistan | 90 | 90 |
| 1 | Albania | 40 | 80 |
| 2 | Angola | 60 | 90 |
| 3 | Antigua and Barbuda | 32 | 64 |
| 4 | Argentina | 60 | 110 |
I think these dataset could help me address the problem by comparing the deaths in each country and see how there is relationship between these rules and the deaths. We can also see whether how does the policy about maximum speed limits and drink-driving affect on the amount of different type of traffic deaths.
We can build models of combination of differnet maximum speed limit and different blood alcohol concentration limit, and predict the possible death; finally find whether there is a best combination that can minimize the traffic death.