DATA100-L5: Data Wrangling and EDA

EDA: Exploratory Data Analysis

Infinite loop

DW & EDA……

DW: raw data -> clean data -> usable data

key data properties to consider in EDA

structure

• file format
  • rectangular data: tables and matrices
  • CSV, TabSV/TSV, json(is a dict)
  • txt, XML

1

    pd.read_csv('filename.tsv',delimiter='\t')

turn to lec jupyter notebook to see more details

• 变量种类 注意：不唯一，不全面
• multiple files 主键？

granularity(颗粒度？) scope and temporality

颗粒度

scope: sampling frame

temporality: time-series data

unix time posix time

faithfulness

missing data?