Creating large Pandas DataFrames: preallocation vs append vs concat

Your benchmark is actually too small to show the real difference.
Appending, copies EACH time, so you are actually doing copying a size N memory space N*(N-1) times. This is horribly inefficient as the size of your dataframe grows. This certainly might not matter in a very small frame. But if you have any real size this matters a lot. This is specifically noted in the docs here, though kind of a small warning.

In [97]: df = DataFrame(np.random.randn(100000,20))

In [98]: df['B'] = 'foo'

In [99]: df['C'] = pd.Timestamp('20130101')

In [103]: df.info()
<class 'pandas.core.frame.DataFrame'>
Int64Index: 100000 entries, 0 to 99999
Data columns (total 22 columns):
0     100000 non-null float64
1     100000 non-null float64
2     100000 non-null float64
3     100000 non-null float64
4     100000 non-null float64
5     100000 non-null float64
6     100000 non-null float64
7     100000 non-null float64
8     100000 non-null float64
9     100000 non-null float64
10    100000 non-null float64
11    100000 non-null float64
12    100000 non-null float64
13    100000 non-null float64
14    100000 non-null float64
15    100000 non-null float64
16    100000 non-null float64
17    100000 non-null float64
18    100000 non-null float64
19    100000 non-null float64
B     100000 non-null object
C     100000 non-null datetime64[ns]
dtypes: datetime64[ns](1), float64(20), object(1)
memory usage: 17.5+ MB

Appending

In [85]: def f1():
   ....:     result = df
   ....:     for i in range(9):
   ....:         result = result.append(df)
   ....:     return result
   ....: 

Concat

In [86]: def f2():
   ....:     result = []
   ....:     for i in range(10):
   ....:         result.append(df)
   ....:     return pd.concat(result)
   ....: 

In [100]: f1().equals(f2())
Out[100]: True

In [101]: %timeit f1()
1 loops, best of 3: 1.66 s per loop

In [102]: %timeit f2()
1 loops, best of 3: 220 ms per loop

Note that I wouldn’t even bother trying to pre-allocate. Its somewhat complicated, especially since you are dealing with multiple dtypes (e.g. you could make a giant frame and simply .loc and it would work). But pd.concat is just dead simple, works reliably, and fast.

And timing of your sizes from above

In [104]: df = DataFrame(np.random.randn(2500,40))

In [105]: %timeit f1()
10 loops, best of 3: 33.1 ms per loop

In [106]: %timeit f2()
100 loops, best of 3: 4.23 ms per loop