class: center, middle, inverse, title-slide # Preliminary Security Analysis from Leblon ### Data Squad ### Gabriel LTDA ### 14/02/20 (updated: 21/02/2020) --- # Summary - Indicators: Definition - Zones: Definition - Trends - Seasonality - Correlation - Next Steps --- # Indicators We have four indicators, named as **Robbery**,**theft**, **Letal Violence** and **Non Letal Violence**. This indicators are the agreggation from the indicators below: 1. Robbery = Business + Residence + vehicle + Pedestrian + Bus + Cell Phone + Driving the victim to withdraw from a financial institution + Bike 1. Theft = vehicle + Pedestrian + Bus + Cell Phone + Bike 1. Lethal Violence = Intentional homicide + Bodily injury plus homicide + Robbery plus homicide (Latrocínio) 1. Non Lethal Violence = Attempted homicide + Bodily injury + Rape --- # Zones We divided in three zones: 1. Rio de Janeiro 1. South Zone (Exclude: Leblon and Ipanema) 1. Leblon and Ipanema --- # Trends In next section, we can see a brief comparative historical about the indicators. --- # Robbery (%) .panelset[ .panel[.panel-name[Rio de Janeiro]
] .panel[.panel-name[Zona Sul]
] ] --- # Highlights - Robbery: Have been **increased** in Leblon and Ipanema, in **contrast** of rest of city. - .can-edit[Your insight] --- # Theft (%) .panelset[ .panel[.panel-name[Rio de Janeiro]
] .panel[.panel-name[South Zone]
] ] --- # Highlights - Theft: Have been **decreased** in Leblon and Ipanema, **following** the rest of city. - .can-edit[Your insight] --- # Lethal Violence (%) .panelset[ .panel[.panel-name[Rio de Janeiro]
] .panel[.panel-name[South Zone]
] ] --- # Highlights - Lethal Violence: Have been **increased** in Leblon and Ipanema, in **contrast** of rest of city. - .can-edit[Your insight] --- # Non Lethal Violence (%) .panelset[ .panel[.panel-name[Rio de Janeiro]
] .panel[.panel-name[South Zone]
] ] --- # Highlights - Non Lethal Violence: Have been **decreased** in Leblon and Ipanema, **following** the rest of city. - .can-edit[Your insight] --- # Seasonality The Leblon and Ipanema presents some seasonality. These Months were indetfied as february and march. Normally, carnaval months. In the nexts plots, the ** blue horizontal line** represent the average from each month. --- # Leblon and Ipanema .panelset[ .panel[.panel-name[Leblon e Ipanema] <!-- --> ] .panel[.panel-name[Zona Sul] <!-- --> ] .panel[.panel-name[Rio de Janeiro] <!-- --> ] ] --- # Highlights - There are some stabilization on index after march, either in theft and robbery. - In contrast with south zone, january is not so danger relative to thefts. - There is a smile-like pattern in the Non-Lethal Violence Index in relation to the South Zone and Rio de Janeiro, but this pattern does not occur in Leblon and Ipanema. - .can-edit[Your insight] --- # Correlation The correlation measuare the strenght of association between two variables. Its represented by a number between -1 and 1. If the correaltion is positive, the variables normally increase or decrease togheter. If be negative, the opposite happens. **Pay attention: Correlation does not imply in causality** In the next slide, you will see a correlation matrix, and each square will contain the number of correlation between the variable in a row and the other variable in a column. If the number is colored with blue, the correlation is positive. If the number is colored with red, the correlation is negative. --- # Disclaimer ## Correlation **classification**: <table> <thead> <tr> <th style="text-align:left;"> Value </th> <th style="text-align:left;"> Classification </th> </tr> </thead> <tbody> <tr> <td style="text-align:left;"> 0 - 0.25 </td> <td style="text-align:left;"> Inexistent </td> </tr> <tr> <td style="text-align:left;"> 0.26 - 0.50 </td> <td style="text-align:left;"> Weak </td> </tr> <tr> <td style="text-align:left;"> 0.51 - 0.75 </td> <td style="text-align:left;"> Moderate </td> </tr> <tr> <td style="text-align:left;"> 0.76 - 1.00 </td> <td style="text-align:left;"> Strong </td> </tr> </tbody> </table> --- # Correlation .panelset[ .panel[.panel-name[Leblon e Ipanema] <!-- --> ] .panel[.panel-name[Zona Sul] <!-- --> ] .panel[.panel-name[Rio de Janeiro] <!-- --> ] ] --- # Highlights - The unique moderate correlation present in the three zones is **theft** and **Non lethal violence **. - In the Leblon and Ipanema, **Roubo** and **Furto** presents a moderate correlation. - There is a strange negative correlation between robbery and Non lethal violence in Rio de Janeiro. --- # Correlation between zones .panelset[ .panel[.panel-name[Roubos] <!-- --> ] .panel[.panel-name[Furtos] <!-- --> ] .panel[.panel-name[Violência Letal] <!-- --> ] .panel[.panel-name[Violência Não Letal] <!-- --> ] ] --- # Highlight - There is a moderate relationship between Leblon and Ipanema and south zone in case of **Robbery**. - There is a almost moderate relationship between Leblon and Ipanema and south zone in case of **theft**. - There is weak relationship between Leblon and Ipanema and south zone in case of **Lethal violence** and **Non lethal violence**. --- # Main highlights >- Robbery and Lethal Violence in Leblon and Ipanema increased too much in last three years, 40% and 175% respectively. >- February and March are the most dangers month of year. The rest of year is stable. >- There`s a moderate correlation between theft and Non lethal violence in all zones. >- There`s a moderate correlation between our neighborhood and the rest of the southern area with regard to theft and robbery. >- .can-edit[Your insight] --- # Next Steps - Find some socioeconomic and demographic data from Leblon. - Find some data more desegregate about the location (like street or lat-long), regards of criminality. - .can-edit[What should i do?]