Data Analysis Toolkit

The tool analyzes numeric columns in a dataset to produce descriptive statistics, detect outliers, and flag data quality issues. It accepts common formats such as CSV or JSON and auto-detects numeric columns. Outputs include per-column summaries (mean, median, std, min, max, count, missing), a list of flagged outliers using the IQR rule, and data-quality flags. Optional analyses include correlation between pairs and a simple linear regression model for target columns. The core logic relies on deterministic, well-known statistical methods and clear handling of missing data. Benefits include faster discovery, reproducible profiling, and a ready-to-share report for analysts, product teams, and researchers. Use cases span financial data checks, A/B test metric review, and scientific data exploration. Unique value comes from multi-column profiling, consistent formatting, and exportable summaries that support downstream analysis without coding. The tool emphasizes correctness and interpretability, providing surfacing of issues like skewness indicators and sample size limitations. It also supports locale-aware number formatting and safe handling of large datasets through chunked processing when applicable. Concepts such as descriptive statistics, outlier detection via 1.5*IQR, correlation coefficients, and regression slope are presented in a compact, actionable form. The result is a structured data quality report that can feed dashboards or data pipelines and improves decision speed for data-driven initiatives.