The official results from the Berlin marathons. The python script scraper.py collects JSON data using the API offered by the Berlin marathon website and writes it in csv format. It was hacked together to get the job done; the code is ugly and uncommented. Although data exists for every year since 1974, the API only offers the data since 2005.
- Names were replaced by their SHA1 hash values
- Bib numbers were removed
- Team names were removed
- The abbreviations for nationalities were replaced by their ISO 3166-1 alpha-3 equivalents as far as possible.
Missing entries were given a value of 'XXX'. - Notation for age class was standardised as describle below.
In particular, entries with no value or a value of 'M' or 'W' were set to 0. - The id fields were removed
There are 9 fields, described below. Some entries are missing for some runners.
| Field | Meaning |
|---|---|
| place | Ranking amongst all runners |
| netTime | Time in seconds measured by runner's chip |
| clockTime | Time in seconds measured from start of race |
| yob | Year of birth |
| ageClass | Age group |
| acPlace | Ranking amongst all runners in the same age group |
| sex | Sex of runner |
| nationality | Nationality of runner |
| name | SHA1 hash value of the name of the runner |
The file abbreviations.csv contains all country codes used. The following abbreviations have not yet been identified:
- GAN
- ISI
- KZK
- MKO
Below are the age groups used in the Berlin marathon.
| Code | Meaning |
|---|---|
| 0 | No age class |
| U20 | Age under 20 |
| 20 | Age between 20 and 30 |
| 30 | Age between 30 and 35 |
| 35 | Age between 35 and 40 |
| 40 | Age between 40 and 45 |
| 45 | Age between 45 and 50 |
| 50 | Age between 50 and 55 |
| 55 | Age between 55 and 60 |
| 60 | Age between 60 and 65 |
| 65 | Age between 65 and 70 |
| 70 | Age between 70 and 75 |
| 75 | Age between 75 and 80 |
| 80 | Age between 80 and 85 |