top of page
  • Writer's pictureAtanu Maity

Building A Data Driven Story: San-Francisco Crime Scenario

Here I have done a detail analysis on San Francisco crime dataset and have looked for some insights from the story hidden behind the data. Find the dataset from here.

Here I have tried to give some visual representation of the underlying story. Most of the graphics are made in Tablue.

So basically here I have looked for some data visualization of the San Francisco crime scenario for the year of 2014-2015.

Across  San Francisco, the Major crime is Larceny/ Theft  which occurred 5,299 times  between the period of Jan,2014 to Jan,2015 followed  by other  offenses(2,807 times), non-criminal (2,717 times) and assault(1,688 times).  So, Larceny/ Theft is the major challenge for the SF police as the graph says. Burglary, Robbery,  Drug are some other crime issues.

It is clear from the picture that southern, northern and central police department districts are most influential crime areas for the criminals. So public safety is under a big threat in these three areas.  In all these three areas Larceny/ Theft is most happening crime event. Police departments of these areas should be more active to catch the thieves.

Across all the PD district also, Larceny/Theft  is the most happening crime. Southern, Northern and Central area are perhaps the best choices for the thieves as the above graph says.

The interesting thing is that among all the theft cases, the major and the most happened is grand theft from locked auto. So clearly road safety is under a big question mark. Also theft from unlocked auto, theft of property, theft from building are some other major theft cases.

But mostly when these theft cases are occurring, is our next point of interest. 

From the above graph, It seems thieves are very active during all over the day. The category- Larceny/ Theft is at the top most position almost every time. 

In Southern district crime rates are comparatively high than other districts. Mission and Northern are also at very high position. Districts with the highest violent crime rates were generally economically disadvantaged localities with high numbers of minorities. Areas with large amounts of tourism also had high crime rates in comparison with other cities.

In the upcoming days, will the scenario be same or there will be any change? Trend analysis can give us some hint.

For  Southern and Northern PD the trend is positive i.e., there is a trend of increasing crime rate in the upcoming days. For Mission, Bayview, Tenderloin and Richmond, the trend is negative, i.e. there is a trend of decreasing crime rate in the upcoming days.

But WHY the performance of the police department is so unsatisfactory!? Hopefully, the below graph gives some clear idea.

Most of the cases has been left out without taking any action/resolution  as NONE has the highest count with a huge majority. But after that 2nd and 3rd highest taken actions/resolutions are ARREST BOOKED and ARREST CITED respectively. 

The important thing is, for the most happening crime i.e for Larcency/Theft, there have been taken NO such steps which the graph shows clearly.

Also the trend for the 3 top most resolutions shows us that NO steps taken cases will increase.


cases it will decrease. So may be in the upcoming days crime cases will increase if there will be no steps taken by the police for public safety.

But if we do an overall trend analysis on crime count, it says something else.

The trend shows crime will decrease over the year as the days go by instead of all these unsatisfactory results and THIS IS VERY INTERESTING!! :-)

#DataScience #DataStory #blogging

30 views0 comments

Recent Posts

See All

Airflow Airflow is a platform to programmatically author, schedule and monitor workflows. Use Airflow to author workflows as Directed Acyclic Graphs (DAGs) of tasks. The main 4 components of Airflow a

bottom of page