Before you upload
- Put column names in the first row (CSV) or use consistent field names (JSON).
- Keep one record per row and one kind of value per column.
- Write dates as
YYYY-MM-DD(for example2025-03-31) so they are recognized as dates. - Stay within the limits: 5 MB per file, 50,000 data rows, 100 columns and 50 datasets per account.
- Remove information you do not need to analyze, especially personal data about other people.
Upload steps
- Open My Datasets in the workspace sidebar.
- Select Upload Dataset and choose a
.csv,.tsv,.txtor.jsonfile. - The file is sent over an encrypted HTTPS connection, validated and parsed on our server, and stored in your account.
- When processing finishes, the dataset opens with a preview of its first records.
Review what was detected
Each column is assigned a type: Numeric, Date, Boolean, Categorical or Text. The type determines which tools can use the column:
- Numeric columns appear in descriptive statistics, histograms, trends, correlations and unusual-value checks.
- Date columns are used for trends and date-range filters.
- Categorical and Boolean columns are used for frequency analysis, group comparisons and cross-tabulations.
If the upload produced warnings — for example, rows with a different number of fields than the header — they are shown above the preview. Read Understanding Supported File Formats to see exactly how types are detected.
Next steps
Check the data quality panel, then continue with Running Descriptive Statistics or Understanding Frequency Distribution.