[python] kaggle, boston marathon

chapter1. pandas다루기_마라톤.ipynb

0.18MB

Ver. Jupyter Notebook (Anaconda3)

▶ 데이터 합치기 : concat

>>> import pandas as pd
>>> import numpy as np

>>> marathon_2015 = pd.read_csv('./date_boston/marathon_results_2015.csv')
>>> marathon_2016 = pd.read_csv('./date_boston/marathon_results_2016.csv')
>>> marathon_2017 = pd.read_csv('./date_boston/marathon_results_2017.csv')


marathon_2015['Year'] = '2015'
marathon_2016['Year'] = '2016'
marathon_2017['Year'] = '2017'

# index 'Offical Time'으로 데이터 합치기
>>> marathon_2015_2017 = pd.concat([marathon_2015, marathon_2016, marathon_2017], ignore_index=True, sort=False)

# 불필요한 columns 삭제
>>> marathon_2015_2017 = marathon_2015_2017.drop(['Unnamed: 0', 'Bib', 'Citizen', 'Unnamed: 9', 'Proj Time', 'Unnamed: 8'], axis=1)
>>> marathon_2015_2017.head()

# for문, pandas로 초단위로 변환
>>> list(marathon_2015_2017.columns)[6:17]

>>> for i in list(marathon_2015_2017.columns)[6:17]:
           marathon_2015_2017[i] = pd.to_timedelta(marathon_2015_2017[i]).astype('m8[s]').astype(np.int64)

>>> print(marathon_2015_2017.shape)
>>> marathon_2015_2017.head()

# 데이터 저장
>>> marathon_2015_2017.to_csv('./data_boston/marathon_2015_2017.csv', index=None, header=True)