如何在python中附加具有相同前缀的多个数据帧

Question

I have multiple sequential dataframe like this:

df1 = pd.DataFrame( [['tom', 10], ['nick', 15], ['juli', 14]] ,  columns = ['Name', 'Age'])
df2 = pd.DataFrame([['tom', 10], ['nick', 15], ['juli', 14]] ,  columns = ['Name', 'Age'])
df3 = pd.DataFrame([['tom', 10], ['nick', 15], ['juli', 14]] ,  columns = ['Name', 'Age'])
df4 = pd.DataFrame([['tom', 10], ['nick', 15], ['juli', 14]] ,  columns = ['Name', 'Age'])

I need to create a for loop to append them and get a new dataframe. I tried the codes below, but it doesnt work, as python recognise df1 as a string.

tempdf = df1
for i in range(2,4):
     tempdf = tempdf.append(("df"+str(i)))
     print(tempdf)

How do I get python to recognise them as dataframe objects I created?

Answer 1

First, I should highlight that having to do this suggests a problem in the way the source dataframes were generated, and you should look into fixing that.

With Python, there are ways to do almost anything you want. Whether it is desirable to make use of such power is another question altogether.

In this case, the safest way would probably be to use globals() :

n_dataframes = 4
g = globals()
dataframes = [g[f'df{i}'] for i in range(1, n_dataframes + 1)]

result_df = pd.concat(dataframes)    
print(result_df)

Output:

   Name  Age
0   tom   10
1  nick   15
2  juli   14
0   tom   10
1  nick   15
2  juli   14
0   tom   10
1  nick   15
2  juli   14
0   tom   10
1  nick   15
2  juli   14

You can perform further processing on the result, such as calling reset_index .

Another alternative is to use eval , which veers firmly into "you shouldn't do this unless you really know what you're doing" territory, because it allows execution of arbitrary code:

dataframes = [eval(f'df{i}') for i in range(1, n_dataframes + 1)]

Note that the above code uses f-strings , which are syntax introduced only in Python 3.6. Accordingly, if your Python version is below that, replace f'df{i}' with 'df{}'.format(i) .

Answer 2

You were proceeding in correct direction, just use eval :

tempdf = df1
for i in range(2,4):
     tempdf = tempdf.append(eval("df"+str(i)))
     print(tempdf)

Note: Using eval can run arbitrary code, using it is considered a bad practice. Please try to use other ways, if possible.