This analysis focuses on a dataset of calls reported to local police agencies concerning domestic violence incidents over the past two decades. Our primary goal is to uncover trends, patterns, and notable anomalies in call volumes, with particular attention to weapon involvement. The project aims to inform future predictive analytics and modeling to better understand and mitigate domestic violence.
Explore call volume distribution over time to identify trends. Analyze weapon involvement in incidents to understand their prevalence. Detect anomalies within call data that may indicate underlying patterns. Establish a foundation for advanced analytics or modeling on domestic violence incidents.
The dataset, DVRCA_2001-2022.csv, encompasses two decades of domestic violence-related call records. Initial data cleaning focused on removing duplicates, handling missing values, and correcting data type inconsistencies to ensure accurate analysis.
Pandas: For data manipulation and cleaning. Matplotlib & Seaborn: For visualizing data trends and distributions. Exploratory Data Analysis (EDA) Our exploratory analysis aimed to dissect the dataset to extract meaningful insights on several fronts:
Yearly Analysis: Revealed a decrease in call volumes from 2001 to 2013, with a slight increase observed until 2017, followed by a decline. Monthly Distribution: Highlighted July as the peak month for calls, potentially attributed to summertime activities.
General Trends: Identified a slightly lower frequency of calls involving weapons compared to those without. Types of Weapons: Firearm-related calls were notably less prevalent than other weapon types, suggesting a need for focused intervention strategies.
Call Distribution: Analysis of calls by county exposed specific regions with disproportionately high volumes, pinpointing areas requiring enhanced policing resources. Agency Focus: Further dissection by agency code offered insights into agencies that field a higher number of calls, guiding resource allocation.
Seasonal Fluctuations: Call volumes exhibit notable seasonal trends, with mid-year peaks. Weapon Involvement: A substantial portion of calls involve weapons, albeit less frequently firearms, emphasizing the need for tailored response strategies. Regional Variations: Certain counties consistently report higher call volumes, highlighting potential hotspots for targeted interventions. Implications for Future Work
Predictive Modeling: Leveraging findings to forecast trends or identify high-risk periods. Policy Impact Assessment: Evaluating the effectiveness of domestic violence policies and interventions. Resource Allocation: Informing strategic decisions on law enforcement and support services distribution.
Stakeholders, including law enforcement agencies, policymakers, and social services, can utilize these insights to formulate strategies aimed at preventing domestic violence and supporting affected individuals more effectively.
Our exploratory data analysis of domestic violence-related calls offers valuable insights into patterns of incidents, weapon involvement, and regional disparities in call volumes. These findings are instrumental in guiding future analytical endeavors, policy formulation, and resource allocation to combat domestic violence more efficiently.