Data Import

Overview

The data-import module reads datasets from third-party formats and transfers them as a new worksheet table into DMAIC.io. It lets you keep working with data produced by other statistics tools without having to retype or roundtrip through CSV.

Supported formats: The Minitab project format .mpx, delimited text files (CSV / TSV / TXT), spreadsheets (.xlsx, .xlsm, .xls, .ods), JSON / NDJSON, and the Q-DAS AQDEF format (.dfq / .dfd / .dfx) are supported today. The module architecture is open — additional formats (JMP, SPSS, …) can be added as further parsers.

Minitab .mpx: An .mpx file is a ZIP archive of XML files that holds a complete Minitab project including all worksheets. DMAIC.io unpacks the archive locally in the browser, reads column names, data types, and values, and makes them available for transfer.

CSV / TSV / TXT: Delimited text files (comma, semicolon, tab, pipe). After loading, delimiter and encoding (UTF-8, Windows-1252, Latin-1) can be switched at any time — the preview is recomputed immediately. Column types (numeric / text) are detected from the values; both period and comma are accepted as decimal separators. For CMM exports (Zeiss CALYPSO, Hexagon PC-DMIS, Mitutoyo MeasurLink) one-click profiles set delimiter, encoding, and the number of metadata rows to skip together — every value remains tweakable afterwards.

Excel / ODS: Spreadsheets from Microsoft Excel (.xlsx, .xlsm, .xls) and LibreOffice / OpenOffice (.ods). All sheets are loaded — use the tabs in the preview to choose which one to transfer. Date cells are converted to ISO format (YYYY-MM-DD); formulas are read as their evaluated values.

JSON / NDJSON: JSON files can appear in three shapes and all of them are auto-detected: a records array (`[{a:1, b:2}, …]` — the worksheet's own export format), a column-oriented object (`{a:[1,2,…], b:[3,4,…]}`), and NDJSON / JSONL with one JSON record per line. Wrapper objects with a `data`, `rows`, or `records` field are unwrapped automatically.

Q-DAS AQDEF: The de-facto SPC / measurement standard in DACH industry (Bosch, Daimler, VW, ZF, Continental). The module reads .dfq, .dfd, and .dfx files and parses the K-keys: K0100 = number of characteristics, K2002/n = characteristic description (column name), K2001/n = characteristic number (fallback name), K0001/n / K0002/n = measurement / attribute values, K1002/n = part description (sheet name). Values may come verbose as individual K0001/n lines or compact with one row per measurement — the row delimiter is detected automatically. Windows-1252 is pre-selected (typical for Q-DAS); UTF-8 with BOM is detected automatically.

Local & offline: The file is unpacked and inspected entirely in the browser — nothing is uploaded to a server. The same holds for sensitive process data.

Operation

• Drag a file onto the drop area or use "Choose file" to pick one through the file dialog.
• The module detects the format from the file extension and unpacks the file immediately.
• The preview shows column names, data types, and the first 20 rows.
• If the file contains several worksheets, switch between the tabs to inspect each one.
• Click "→ Worksheet" to create a new worksheet in the data section from the currently shown table.

Typical use cases: continue working with existing Minitab training datasets in DMAIC.io, load example data from textbooks, reproduce audit reports from third-party systems.

Pitfalls

Missing values: Minitab marks missing values with an asterisk (*). On import these are converted to real empty cells so that statistics modules ignore them correctly.

Date and time columns: Date / time columns are currently imported as text, not as native date values. For time-series analyses you may want to add a helper numeric column.

Pick the right sheet first: For files with multiple sheets, only the currently active sheet is transferred. If you need other sheets, run the module again or import the file repeatedly with a different tab selected each time.

Very large files: Because unpacking happens in the browser, files in the hundreds of megabytes range can take a moment. The module shows "Reading file…" during that time — please wait.

Re-pick after reload: After a browser reload the file name is kept, but the content is not — browsers do not allow persisting file handles. The file has to be picked again in that case.

CSV: check encoding & delimiter: If umlauts or accented characters look garbled in the preview, "Windows-1252 (Excel)" is usually the right encoding. In Germany, Excel often exports with a semicolon instead of a comma — if columns look wrong, switch the delimiter. Either change triggers an immediate re-parse.