Skip to content

Latest commit

 

History

History
22 lines (20 loc) · 1.38 KB

LEX_notes.md

File metadata and controls

22 lines (20 loc) · 1.38 KB
Raw

Notes for users of the LEX matrices

  • The geographic identifiers accompanying the state-level LEX are two-letter postal abbreviations.
  • The geographic identifiers accompanying the county-level LEX are five-digit FIPS county codes.
  • Please remember that these are not transition matrices because a phone can visit multiple counties in a 14-day period. The columns do not sum to one.
  • We report the LEX data as square matrices in CSV files. If you want to reshape them to be "long" (N2 rows with two identifier columns and one column of data), this can be done quickly. Stata's "reshape" command will not do this quickly for a 2000-by-2000 matrix. Try something like this Python script:
import pandas as pd
url = 'https://github.com/COVIDExposureIndices/COVIDExposureIndices/blob/master/lex_data/county_lex_2020-01-20.csv.gz?raw=true'
df 	= pd.read_csv(url, compression='gzip', header=0)
countys = df.columns.values[1:]
col_names = dict(zip(countys, ["a" + lab for lab in countys]  ))
df 	= df.rename(columns = col_names)
df_long	= pd.wide_to_long(df, stubnames="a", i=['COUNTY_PRE'], j='col')
df_long = df_long.reset_index(drop=False)
df_long = df_long.rename(columns = {"col" : "COUNTY", "a" : "LEX"})
df_long.to_csv(r'reshaped.csv',index=False)