用相似的字符串连接两列上的两个数据框 Python

Question

I have two dataframes (df1 and df2) that I want to left merge on using two columns,'State' (ie Arkansas) and 'County' (ie Union). (Union is a county in Arkansas).

df1 and df2 need to match across for 'State' and 'County', but df2 has county names with additional strings (ie Woodmont County Borough) not found in df1 county names (ie Woodmont).

What can I do to left merge these two dataframes with different representations of counties? I have many county names.

Answer 1

First, get a list of the 'County' in df1

Then, create a new column in df2 which, if a county in county_list is found in df2.County, stores it in that new column which we will call County_cleaned

Then for each County in county_list , if it appears in df2['County'] , then place that in the newly created County_cleaned

And now, you can merge df1 and df2 together (lets call it df3), using this newly created column in df2:

# get a list of the counties in df1
county_list = df1.County.unique()

#initialise a new column to empty string
df2['County_cleaned'] = ''

#for each of the counties in df1, if a county from df1 appears 
#somewhere in the df2.County, then add it to the newly created 
#column called County_cleaned 
for c in county_list:
    df2.loc[df2['County'].str.contains(c), 'County_cleaned']=c

#merge the 2 dataframes to create df3
df3 = df1.merge(df, how='inner', left_on=['State','County'], right_on=['State', 'County_cleaned')

Note: I've set how='inner' but this can also be 'outer','left','right' depending on the type of join.