# Caltech CS 155 2019 Part 2 — Data

来源：https://www.kaggle.com/competitions/caltech-cs-155-2019-part-2/data

本地 Meta Kaggle 快照；不是本次实时页面。

# Dataset Description

The data provided in and `test_2012.csv` primarily follows the documentation provided in the file `2012-Codebook.pdf`. The following fields in the codebook have been removed and thus not in the `.csv` files:

`['HRHHID', 'PROLDRRP', 'PUPELIG', 'HXFAMINC', 'PUSCK4', 'PES1', 'PES2', 'PES3', 'PES4', 'PES5', 'PES6', 'PES7', 'PES8']`

The field `HRHHID` has been renamed as `id` in the `.csv` files, which starts from 0 and uniquely identifies each row in the dataset.

### Task

Your task is to predict the `target` column. `target` being 0  the person did not vote in the election; if `target` is 1, the person voted in the election. (In fact, this `target` column directly corresponds to the field `PES1` in the codebook. Except that `PES1` has values 1 and 2, and we shifted it to 0 and 1.)

Please note that your submission should be **_the probabilities_** of each testing sample being `1` (i.e., floating point numbers), instead of binary 0/1.

### Features

The attached PDF codebook explains what each feature represents.

### File descriptions

- test_2012.csv: the test set, with no labels
- sample_submission_2012.csv: an example submission showing the format required
- 2012-Codebook.pdf: explanation of the various features