Is there a way to round a single column in pandas without affecting the rest of the dataframe?

>>> print(df)
  item  value1  value2
0    a    1.12     1.3
1    a    1.50     2.5
2    a    0.10     0.0
3    b    3.30    -1.0
4    b    4.80    -1.0

I have tried the following:

>>> df.value1.apply(np.round)
0    1
1    2
2    0
3    3
4    5
5    5

What is the correct way to make data look like this:

  item  value1  value2
0    a       1     1.3
1    a       2     2.5
2    a       0     0.0
3    b       3    -1.0
4    b       5    -1.0
5    c       5     5.0

You are very close.
You applied the round to the series of values given by df.value1.
The return type is thus a Series.
You need to assign that series back to the dataframe (or another dataframe with the same Index).

Also, there is a pandas.Series.round method which is basically a short hand for pandas.Series.apply(np.round).

>>> df.value1 = df.value1.round()
>>> print df
  item  value1  value2
0    a       1     1.3
1    a       2     2.5
2    a       0     0.0
3    b       3    -1.0
4    b       5    -1.0

For some reason the round() method doesn’t work if you have float numbers with many decimal places, but this will.

decimals = 2    
df['column'] = df['column'].apply(lambda x: round(x, decimals))

Use the pandas.DataFrame.round() method like this:

df = df.round({'value1': 0})

Any columns not included will be left as is.

No need to use for loop. It can be directly applied to a column of a dataframe

sleepstudy['Reaction'] = sleepstudy['Reaction'].round(1)

If you are doing machine learning and use tensorflow, many float are of ‘float32’, not ‘float64’, and none of the methods mentioned in this thread likely to work. You will have to first convert to float64 first.

x.astype('float')

before round(…).

No need to use lambda or creating function. It is straight-forward.
See example below

df['decimal_place_2'] = df['decimal_place_2'].round(2)

saldo_acred['cumsum_prc_saldo'].astype('float').round(2)

In my case, I have both string values as well as decimal values in single columns.

def round_2(x):
    try:
        return round(x,2)
    except:
        return x

df['cur_TMIN_IMD_WeekSum']=df['cur_TMIN_IMD_WeekSum'].apply(round_2)