How to preview a part of a large pandas DataFrame, in iPython notebook?

df.head(5) # will print out the first 5 rows
df.tail(5) # will print out the 5 last rows

In this case, where the DataFrame is long but not too wide, you can simply slice it:

>>> df = pd.DataFrame({"A": range(1000), "B": range(1000)})
>>> df
<class 'pandas.core.frame.DataFrame'>
Int64Index: 1000 entries, 0 to 999
Data columns:
A    1000  non-null values
B    1000  non-null values
dtypes: int64(2)
>>> df[:5]
   A  B
0  0  0
1  1  1
2  2  2
3  3  3
4  4  4

ix is deprecated.

If it's both wide and long, I tend to use .ix:

>>> df = pd.DataFrame({i: range(1000) for i in range(100)})
>>> df.ix[:5, :10]
   0   1   2   3   4   5   6   7   8   9   10
0   0   0   0   0   0   0   0   0   0   0   0
1   1   1   1   1   1   1   1   1   1   1   1
2   2   2   2   2   2   2   2   2   2   2   2
3   3   3   3   3   3   3   3   3   3   3   3
4   4   4   4   4   4   4   4   4   4   4   4
5   5   5   5   5   5   5   5   5   5   5   5

I write a method to show the four corners of the data and monkey-patch to dataframe to do so:

def _sw(df, up_rows=10, down_rows=5, left_cols=4, right_cols=3, return_df=False):
    ''' display df data at four corners
        A,B (up_pt)
        C,D (down_pt)
        parameters : up_rows=10, down_rows=5, left_cols=4, right_cols=3
        usage:
            df = pd.DataFrame(np.random.randn(20,10), columns=list('ABCDEFGHIJKLMN')[0:10])
            df.sw(5,2,3,2)
            df1 = df.set_index(['A','B'], drop=True, inplace=False)
            df1.sw(5,2,3,2)
    '''
    #pd.set_printoptions(max_columns = 80, max_rows = 40)
    ncol, nrow = len(df.columns), len(df)

    # handle columns
    if ncol <= (left_cols + right_cols) :
        up_pt = df.ix[0:up_rows, :]         # screen width can contain all columns
        down_pt = df.ix[-down_rows:, :]
    else:                                   # screen width can not contain all columns
        pt_a = df.ix[0:up_rows,  0:left_cols]
        pt_b = df.ix[0:up_rows,  -right_cols:]
        pt_c = df[-down_rows:].ix[:,0:left_cols]
        pt_d = df[-down_rows:].ix[:,-right_cols:]

        up_pt   = pt_a.join(pt_b, how='inner')
        down_pt = pt_c.join(pt_d, how='inner')
        up_pt.insert(left_cols, '..', '..')
        down_pt.insert(left_cols, '..', '..')

    overlap_qty = len(up_pt) + len(down_pt) - len(df)
    down_pt = down_pt.drop(down_pt.index[range(overlap_qty)]) # remove overlap rows

    dt_str_list = down_pt.to_string().split('\n') # transfer down_pt to string list

    # Display up part data
    print up_pt

    start_row = (1 if df.index.names[0] is None else 2) # start from 1 if without index

    # Display omit line if screen height is not enought to display all rows
    if overlap_qty < 0:
        print "." * len(dt_str_list[start_row])

    # Display down part data row by row
    for line in dt_str_list[start_row:]:
        print line

    # Display foot note
    print "\n"
    print "Index :",df.index.names
    print "Column:",",".join(list(df.columns.values))
    print "row: %d    col: %d"%(len(df), len(df.columns))
    print "\n"

    return (df if return_df else None)
DataFrame.sw = _sw  #add a method to DataFrame class

Here is the sample:

>>> df = pd.DataFrame(np.random.randn(20,10), columns=list('ABCDEFGHIJKLMN')[0:10])

>>> df.sw()
         A       B       C       D  ..       H       I       J
0  -0.8166  0.0102  0.0215 -0.0307  .. -0.0820  1.2727  0.6395
1   1.0659 -1.0102 -1.3960  0.4700  ..  1.0999  1.1222 -1.2476
2   0.4347  1.5423  0.5710 -0.5439  ..  0.2491 -0.0725  2.0645
3  -1.5952 -1.4959  2.2697 -1.1004  .. -1.9614  0.6488 -0.6190
4  -1.4426 -0.8622  0.0942 -0.1977  .. -0.7802 -1.1774  1.9682
5   1.2526 -0.2694  0.4841 -0.7568  ..  0.2481  0.3608 -0.7342
6   0.2108  2.5181  1.3631  0.4375  .. -0.1266  1.0572  0.3654
7  -1.0617 -0.4743 -1.7399 -1.4123  .. -1.0398 -1.4703 -0.9466
8  -0.5682 -1.3323 -0.6992  1.7737  ..  0.6152  0.9269  2.1854
9   0.2361  0.4873 -1.1278 -0.2251  ..  1.4232  2.1212  2.9180
10  2.0034  0.5454 -2.6337  0.1556  ..  0.0016 -1.6128 -0.8093
..............................................................
15  1.4091  0.3540 -1.3498 -1.0490  ..  0.9328  0.3668  1.3948
16  0.4528 -0.3183  0.4308 -0.1818  ..  0.1295  1.2268  0.1365
17 -0.7093  1.3991  0.9501  2.1227  .. -1.5296  1.1908  0.0318
18  1.7101  0.5962  0.8948  1.5606  .. -0.6862  0.9558 -0.5514
19  1.0329 -1.2308 -0.6896 -0.5112  ..  0.2719  1.1478 -0.1459


Index : [None]
Column: A,B,C,D,E,F,G,H,I,J
row: 20    col: 10


>>> df.sw(4,2,3,4)
        A       B       C  ..       G       H       I       J
0 -0.8166  0.0102  0.0215  ..  0.3671 -0.0820  1.2727  0.6395
1  1.0659 -1.0102 -1.3960  ..  1.0984  1.0999  1.1222 -1.2476
2  0.4347  1.5423  0.5710  ..  1.6675  0.2491 -0.0725  2.0645
3 -1.5952 -1.4959  2.2697  ..  0.4856 -1.9614  0.6488 -0.6190
4 -1.4426 -0.8622  0.0942  .. -0.0947 -0.7802 -1.1774  1.9682
..............................................................
18  1.7101  0.5962  0.8948  .. -0.8592 -0.6862  0.9558 -0.5514
19  1.0329 -1.2308 -0.6896  .. -0.3954  0.2719  1.1478 -0.1459


Index : [None]
Column: A,B,C,D,E,F,G,H,I,J
row: 20    col: 10