Skip to content

Data Import

The first tab loads your data file and maps its columns to the standard names the rest of the pipeline expects.

Loading a file

Click Load Data and choose a file. The following formats are supported:

  • CSV (.csv)
  • Excel (.xlsx, .xls)

The first few rows are shown in a preview table so you can confirm the contents.

Column mapping

Because source files use different column names, you map them to the standard fields:

Standard field Required Meaning
date :material-check: Observation date/timestamp.
price :material-check: Observed price.
product_id :material-check: Product/item identifier.
quantity optional Quantity sold (enables weighted methods).

Each field has a dropdown populated with your file's actual column names. Select None for the optional quantity column if your data does not contain quantities.

Date format

Provide the strptime pattern matching your date column. The default is:

%Y-%m-%d

Common alternatives:

Format string Example match
%d/%m/%Y 15/06/2026
%m/%d/%Y 06/15/2026
%Y%m%d 20260615

Standardize

Click Standardize Columns. The app validates and converts the mapped columns to the expected types and shows a summary:

  • number of products,
  • the date range of the data,
  • whether quantity data is available.

Weighted methods need quantity

If your file has no quantity column, weighted aggregation and weighted index methods (Laspeyres, Paasche, Fisher, Törnqvist, Walsh, GEKS-Fisher, GEKS-Törnqvist, Geary-Khamis) will be unavailable downstream.

Use one row per product per observation date. For example:

date product_id price quantity
2026-01-01 A 100 10
2026-01-01 B 200 20
2026-02-01 A 110 12
2026-02-01 B 190 18

Once the summary appears, continue to Aggregation.