# Bidding Predictions for Construction — Data

来源：https://www.kaggle.com/competitions/bidding-predictions-for-construction/data

采集时间：2026-10-19T10:48:43Z

Data includes publicly available bid results. Bid results are provided for each competitor and line item. Bid results are provided at the contractor–job–pay-item level. These may be aggregated to compute each contractor’s total bid for a given job, which is the unit used for evaluation. All job types and sizes will be included in the testing dataset. Participants may choose to model directly from the job-level datasets or engineer features by aggregating pay-item–level data; however, all submissions are evaluated at the contractor–job level only.

## Files

*   **raw_train.csv** - the pay-item level training set, there are multiple pay-items that make up each job total bid
*   **train_summary.csv** - the training set in a by-job level view that mirrors how test is evaluated
*   **raw_test.csv** - pay-item level predictors for the test set
*   **test.csv** - the test set in a contractor–job–level view that mirrors how submissions are evaluated with appropriate job-level predictors
*   **sample_submission.csv** - a sample submission file in the correct format

## Columns

*   `job_id` - str - unique job identifier that can have multiple contractor_id bidders and pay_item_ids that make up the total bid
*   `contractor_id` - str - unique contractor identifier that can be used to determine by-contractor bidding strategies
*   `bid_date` - date - date the job was bid on
*   `total_bid` - f64 - the total bid submitted, a summation of individual pay items.
*   `pay_item_id` - str - unique pay-item identifier
*   `pay_item_description` - str - description of what each pay-item is
*   `unit_english_id` - str - clarification of what unit each pay-item is measured in
*   `quantity` - f64 - the total amount of occurrences of that pay-item
*   `amount` - f64 - the UNIT PRICE submitted for that pay-item
*   `job_category_description` - str - category the job fits into
*   `category_id` - i32 - unique category identifier
*   `primary_location` - str - location of the job
*   `row_id` - str - unique identifier representing a single contractor’s bid on a single job (job_id + contractor_id), used to align submissions with the solution for scoring.