36

Say I have an dictionary of dataframes:

  {'df1': name         color    type
          Apple        Yellow   Fruit,
   'df2': name         color    type
          Banana       Red      Fruit,
   'df3': name         color    type
          Chocolate    Brown    Sweet
    ......}

And I want to merge them all into one like this:

  name         color    type
  Apple        Red      Fruit 
  Banana       Yellow   Fruit
  Chocolate    Brown    Sweet

I can do it manually as follows:

  merge1=pd.merge('df1','df2')
  merge2=pd.merge('merge1','df3')
  ...

But is there a way to automatically zip through the dictionary and merge? Any help is appreciated.

qts
  • 984
  • 2
  • 14
  • 25

2 Answers2

67

You can just pass the dict direct and access the values attribute to concat:

In [233]:

d
Out[233]:
{'df1':     name         color    type
 0  Apple        Yellow   Fruit, 'df2':     name         color    type
 0  Banana       Red      Fruit, 'df3':     name         color    type
 0  Chocolate    Brown    Sweet}
In [234]:

pd.concat(d.values(), ignore_index=True)
Out[234]:
    name         color    type
0  Banana       Red      Fruit
1  Apple        Yellow   Fruit
2  Chocolate    Brown    Sweet

This assumes that you are just looking to concatenate all the dfs, if you are going to merge then you need to explain what the merge criteria is

EdChum
  • 376,765
  • 198
  • 813
  • 562
3

You can concatenate all dataframes in a dictionary by doing this

print(d.keys())

df_names = []
for i in d.keys():
    temp_df = d[i]
    df_names.append(temp_df)

d['merged'] = pd.concat(df_names)
Fiona Victoria
  • 41
  • 3