比较两个数据框，找到共同的元素，如果不存在则填充列值

Question

So I have two data frames, with some common keywords.

for example:

df1 = {'keyword': ['Computer','Phone','Printer'],
       'Price1':   [1200,800,200],
       'category':['first','second','first']
       }


df2= {'keyword': ['Computer','Phone','Printer','chair'],
      'Price2': [1200,800,200,40]
      }

As you can see above, one df has a category feature, while the other doesn't. So what I want to do is combine two dfs, keep the common items as it is, and if there are some keywords present in one df ('chair', in our case), and absent in another, add the values from df where that keyword exists,and fill that categorical feature (category) with a particular value with 'third' for example.

Answer 1

While not entirely clear, I think you want combine_first :

df2.combine_first(df1)

NB. I transformed the dictionaries to dataframes first with dfX = pd.DataFrame(dfX)

output:

   Price1  Price2 category   keyword
0  1200.0    1200    first  Computer
1   800.0     800   second     Phone
2   200.0     200    first   Printer
3     NaN      40      NaN     chair

Alternatively, use merge :

df1.merge(df2, on='keyword', how='outer')

output:

    keyword  Price1 category  Price2
0  Computer  1200.0    first    1200
1     Phone   800.0   second     800
2   Printer   200.0    first     200
3     chair     NaN      NaN      40

Answer 2

Building upon mozway's answer, if the prices of the items do not vary across the DataFrames you don't need to specify Price1 and Price2 in the column names. Also, after joining the data, you can fill the remaining NAs in the Category column with any word you want with the fillna() .

Here is the streamlined code for you:

import pandas as pd


df1 = pd.DataFrame({'keyword': ['Computer','Phone','Printer'],
       'Price':   [1200,800,200],
       'category':['first','second','first']
       })


df2 = pd.DataFrame({'keyword': ['Computer','Phone','Printer','chair'],
      'Price': [1200,800,200,40]
      })

df_combined = df1.combine_first(df2)

# Arbitrarily sets the word for unknown categories
keyword = "third"

df_combined["category"].fillna(keyword, inplace=True)

And this is its output:

    Price category   keyword
0  1200.0    first  Computer
1   800.0   second     Phone
2   200.0    first   Printer
3    40.0    third     chair