CODEWITHSUNDEEP
pythonpandas
Brandon
Brandon

Reputation: 7966

Converting pandas column to string w/o scientific notation

Lots of questions address this, but none of the solutions seem to work exactly as I need.

I have a dataframe with two columns of numbers with 10-20 digits each. These are actually ID #s, and I'd like to concatenate them. It looks like that's best done by first converting the values to strings.

However, when converting with .astype(str), pandas keeps the scientific notation, which won't fly.

Things I've tried:

tried: dtype arg ('str') or converters (using str()) in read_csv()

outcome: df.dtypes still lists 'objects,' and values still display in sci. notation

tried: pd.set_option('display.float_format', lambda x: '%.0f' % x)

outcome: displays good in df.head(), but reverts to scientific notation upon coercion to string & concatenation using + operator

tried: coercing to int, str, or str(int(x)).
outcome: int works when i coerce one value with int(), but not when I use astype(int). using .apply() with int() throws an 'invalid literal long() with base 10' error.

This feels like it should be pretty straightforward, anxious to figure out what I'm missing.

Upvotes: 2

Views: 2104

Answers (1)

ako
ako

Reputation: 3689

What you tried sets the display format. You could just format the float as a string in the dataframe.

import numpy as np
import pandas as pd

import numpy as np

import pandas as pd

df=pd.DataFrame(data={'a':np.random.randint(low=1,high=100,size=10)*1e20,'b':np.random.randint(low=1,high=100,size=10)*1e20})
df.apply(lambda x: '{0:20.0f}|{1:20.0f}'.format(x.a,x.b),axis=1)



Out[34]:
0    9699999999999998951424|4600000000000000000000
1     300000000000000000000|2800000000000000000000
2    9400000000000000000000|9000000000000000000000
3    2100000000000000000000|4500000000000000000000
4    5900000000000000000000|4800000000000000000000
5    7700000000000000000000|6200000000000000000000
6    1600000000000000000000|8000000000000000000000
7      100000000000000000000|400000000000000000000
8    9699999999999998951424|8000000000000000000000
9    4500000000000000000000|3500000000000000000000

Upvotes: 2

Related Questions