MAG OptiAI

Operational AI Input Specs

Forecasting Input Specs

Use this guide to prepare the CSV accepted by the Forecasting workspace. Every upload needs a timestamp column plus a numeric target column, and the workspace supports both single-variate uploads and multivariate uploads with additional numeric driver columns.

Overview

Use this guide to prepare valid input files for this workspace. Review the required files, exact column names, example rows, and common formatting rules before uploading.

Required files

2

Optional files

0

Spec sections

2

Applies to single-variate

Single-variate Forecast CSV

Use this shape when the scenario is driven only by the historical target series. The CSV needs one timestamp column and one numeric target column.

Required file

forecast-series.csv

Single time-series dataset used for Forecasting lifecycle training and forecast runs.

Requirements

Required / alternative columns

  • date or ds
  • y

Optional columns

  • numeric driver columns such as ad_spend, promo_index, price_index

Approved aliases

  • date: ds

Format limitations

  • Timestamp values currently use YYYY-MM-DD or ISO datetime formats.

Unsupported currently

  • Nonnumeric driver columns are not used as model features; they are warned and ignored.

Value rules and examples

  • Include either a date column or a ds column plus a numeric y column.
  • Current timestamp policy accepts YYYY-MM-DD or ISO datetime values; broader date parsing is not part of this contract.
  • Nonnumeric driver columns are surfaced as warnings and ignored rather than used as model features.
  • Any extra columns should be numeric when you want driver-aware runs to use them.

Example CSV table

dateyad_spendpromo_index
2024-01-0141018.21.04
2024-01-0241818.61.03

Common mistakes

  • Leaving out the date or ds column.
  • Uploading non-numeric values in y.
  • Assuming the workspace needs extra columns for a basic univariate Forecasting run.

Quick rules

  • Every Forecasting upload needs either date or ds together with y.
  • Single-variate uploads should stop at the timestamp and target columns.
  • Use clean ISO-style timestamps when possible so the workspace can infer frequency reliably.

Navigate

Keep the right guide close by

Pair this upload guide with the product page for workflow context or the resources hub for related public material.

Applies to multivariate

Multivariate Forecast CSV

Use this shape when you want Forecasting to score the target together with additional numeric driver columns such as spend, price, or demand drivers.

Required file

forecast-series.csv

Single time-series dataset used for Forecasting lifecycle training and forecast runs.

Requirements

Required / alternative columns

  • date or ds
  • y

Optional columns

  • numeric driver columns such as ad_spend, promo_index, price_index

Approved aliases

  • date: ds

Format limitations

  • Timestamp values currently use YYYY-MM-DD or ISO datetime formats.

Unsupported currently

  • Nonnumeric driver columns are not used as model features; they are warned and ignored.

Value rules and examples

  • Include either a date column or a ds column plus a numeric y column.
  • Current timestamp policy accepts YYYY-MM-DD or ISO datetime values; broader date parsing is not part of this contract.
  • Nonnumeric driver columns are surfaced as warnings and ignored rather than used as model features.
  • Any extra columns should be numeric when you want driver-aware runs to use them.

Example CSV table

dateyad_spendpromo_index
2024-01-0141018.21.04
2024-01-0241818.61.03

Common mistakes

  • Using text labels in driver columns you expect the workspace to treat as numeric features.
  • Dropping the base date or ds and y columns while adding drivers.
  • Assuming extra columns are required for every Forecasting upload.

Quick rules

  • Multivariate uploads still need the same base timestamp and y columns as every Forecasting run.
  • Additional driver columns should stay numeric so the workspace can treat them as valid multivariate inputs.
  • Classical univariate methods and driver-aware runs do not rely on the same shape, so keep the driver columns only when you need them.

Navigate

Keep the right guide close by

Pair this upload guide with the product page for workflow context or the resources hub for related public material.

Navigate

Keep the right input guide close

Return to the input-spec library for another upload contract, or go back to the resource hub for quickstarts, how-to guides, and use cases.