WORKSPACE GUIDE
Understand your report
A few things to know before you explore your dataset.
Prepare your CSV
Choose a comma, semicolon, tab, or pipe delimiter. By default, the first row supplies column names. Select “No” if the file has no header; column names will be assigned automatically.
Empty cells and standard missing markers such as ?, NA, N/A, and NULL are recognized automatically. Add custom markers separated by commas; these markers are case-sensitive.
Read the quality overview
“Completed” means the analysis finished successfully. It does not mean the dataset has no issues. The issue total counts each column with missing values, each numeric column with outliers, one finding if duplicate rows exist, and each structural warning.
Explore missing values
Percentages use the total number of dataset rows. Severity is none at 0%, low below 5%, moderate from 5% to below 20%, and high at 20% or more. An unavailable value is different from zero.
Review outliers
The IQR check flags values below Q1 − 1.5 × IQR or above Q3 + 1.5 × IQR. Values exactly on a bound are included in the expected range. Percentages use the non-missing values in each numeric column.
Outliers are observations to review, not automatically incorrect data. Use your knowledge of the dataset before changing or removing values.
Saved reports
Save a report’s URL to revisit it. Reports contain metadata and diagnostic summaries; raw uploaded rows are not retained. Structural warnings are shown as findings and do not automatically mean the application failed.
Create an account to analyze your data