India Suicide Statistics Analysis (2001-2012)

Solution & Approach

• Multi-Dimensional Dataset Preparation: Loaded CSV dataset containing 237,519 suicide records spanning 2001-2012 with 7 variables (State, Year, Type_code, Type, Gender, Age_group, Total). Performed data cleaning by removing aggregate categories ('Total (All India)', 'Total (States)', 'Total (Uts)') and filtering out broad age group '0-100+' to prevent double-counting. Final cleaned dataset contained 225,663 granular records across 35 states, 12 years, 56 suicide types, 2 genders, and 5 distinct age groups
• Temporal Trend Analysis: Created year-wise aggregation using Pandas groupby() to calculate total suicide counts from 2001-2012. Generated line plot with Matplotlib revealing significant upward trend from 325,428 cases (2001) to peak of 405,989 (2011), followed by decline to 376,398 (2012). Applied time series visualization with custom axis formatting using matplotlib.ticker for readability. Identified 2011 as critical inflection point requiring policy attention
• Gender-Based Statistical Comparison: Built pivot table with Year as index and Gender as columns to compare male vs female suicide rates over time. Analysis revealed persistent 2:1 gender disparity with male suicides ranging from 198,935 (2001) to 262,777 (2011) compared to female rates of 126,493 to 143,212. Prepared independent samples t-test using scipy.stats.ttest_ind to validate statistical significance of gender as predictor variable, testing null hypothesis that gender has no effect on suicide rates
• Geographic Hotspot Identification: Performed state-wise aggregation to identify regions with highest absolute suicide counts. Created grouped bar charts using Seaborn to visualize state distributions, highlighting Maharashtra, Madhya Pradesh, and Karnataka as top three states. Implemented data sorting and filtering to isolate top 10 states for focused visualization, enabling targeted state-level policy recommendations
• Age Group Segmentation Analysis: Conducted age-stratified analysis across 5 groups to identify vulnerable demographics. Used Pandas crosstab functionality to create age_group vs gender contingency tables. Visualized age distribution patterns with grouped bar plots to reveal that working-age populations (15-44 years) account for majority of cases, suggesting economic stress as potential contributing factor
• Causal Factor Classification: Analyzed Type variable containing 56 distinct suicide causes including economic stress, dowry disputes, illness, unemployment, and family problems. Created frequency distribution charts ranking causes by prevalence. Implemented multi-label categorization to group related causes (economic, social, health-related) for higher-level trend analysis. Generated Seaborn heatmaps showing correlation between cause types and demographic variables
• Interactive HTML Visualization Export: Converted Jupyter notebook to standalone HTML format using nbconvert for stakeholder presentation. Generated publication-ready visualizations with consistent color schemes, proper axis labels, and statistical annotations. Created comprehensive analysis report PDF documenting methodology, findings, hypothesis test results, and policy recommendations for public health officials