Methods Lab · Data Check
Understand the quality of your CSV.
Review missing values, types, duplicates and explicit rules. Keep the original file and download a descriptive report.
The first opening downloads the R engine and packages. This executable environment has separate security settings; no R installation is required.
Start with the synthetic example
The app opens with 121 fictitious visits and a data dictionary. Confirm reading options and types before checking another file.
What it checks
Declared missing tokens, conversion failures, repeated rows and identifiers, distributions, IQR candidates and user-declared rules. Repeated identifiers may represent valid visits; outlier candidates require context.
Limits and interpretation
Maximum 5 MiB, 20,000 rows, 100 columns and 500,000 cells. Each row is a file record: declare its unit and any rule units. No automatic cleaning, imputation, models or clinical quality certification.
Local file reading and report download
The selector reads CSV bytes into browser memory. The app does not upload the selected CSV to a server. The downloaded report may contain file names, categories and values; handle it according to their sensitivity. Closing or reloading discards the session.
Use the synthetic example or public, non-identifiable data with explicit provenance. Do not enter patient-level, confidential or identifiable data.