how to convert df.info() into data frame. df.info()

user_14341061 0
Viewed 1272

how to convert df.info() into data_frame.

I want to be able to merge this dataframe with other dataframe.

df.info()
<class 'pandas.core.frame.DataFrame'>
RangeIndex: 458 entries, 0 to 457
Data columns (total 9 columns):
Name        457 non-null object
Team        457 non-null object
Number      457 non-null float64
Position    457 non-null object
Age         457 non-null float64
Height      457 non-null object
Weight      457 non-null float64
College     373 non-null object
Salary      446 non-null float64
dtypes: float64(4), object(5)
memory usage: 32.3+ KB
pandas dataframe converters merge python-3.x
3 Answers

While the info() method directly prints the information to the output, and even when you use the buffer to extract the info, it extracts it in the form of text lines, which are hardly useful for further processing.

The above mentioned solution does work, but it creates a problem when your column names have spaces or are inconsistently named for you to be able to use the line.split() with some other separator character.

I couldn't search of any way to do this using the default info() method. So, I made my own function to do this. And it is not that complicated.

def infoOut(data,details=False):
    dfInfo = data.columns.to_frame(name='Column')
    dfInfo['Non-Null Count'] = data.notna().sum()
    dfInfo['Dtype'] = data.dtypes
    dfInfo.reset_index(drop=True,inplace=True)
    if details:
        rangeIndex = (dfInfo['Non-Null Count'].min(),dfInfo['Non-Null Count'].min())
        totalColumns = dfInfo['Column'].count()
        dtypesCount = dfInfo['Dtype'].value_counts()
        totalMemory = dfInfo.memory_usage().sum()
        return dfInfo, rangeIndex, totalColumns, dtypesCount, totalMemory
    else:
        return dfInfo

Usage:

variable = infoOut(yourDataFrameObject)
#or
var1, var2, var3, var4, var5 = infoOut(yourDataFrameObject,details=True)

This function will return the exact table structure as returned by the info method for your supplied dataframe, and that too in a dataframe format. Further, if you supply with an argument details=True, then it will also give our other information that info() gives out like memory, summary counts, etc.

Modify the function as you like.

Good day.

user_109161360

You can write to a text file :

import io
buffer = io.StringIO()
df.info(buf=buffer)
i = buffer.getvalue()
with open("df_info.txt", "w", encoding="utf-8") as f:
    f.write(i)
user_98513930

Building off of the previous answer. The solution below will place the string collected from the buffer directly into a pandas DataFrame without having to save a temp file to disk

import io
buf = io.StringIO()
df.info(buf=buf)
s = buf.getvalue()
lines = [line.split() for line in s.splitlines()[3:-2]]
pd.DataFrame(lines)

Explaination:

  • s.splitlines() - creates a list from the string where a new line character is found
  • indexing [3:-2] - removes the first three lines and last two so that it will fit nicely into columns for the data frame
user_24025010
Related