Python dataframe logic

Language Help Python
1

Hi, I want to ask regarding python index and Data Frame

Given DataFrame:
   A    B    C
0  1  NaN    3
1  4    5  NaN
2  7    8    9
 
if we do this: diabetes_data.isnull()
it will:

       A      B      C
0  False   True  False
1  False  False   True
2  False  False  False

However, this is the part that I dont understand when we index it to the original DF.
diabetes_data[diabetes_data.isnull()]

 
A      B      C
0  NaN   NaN    NaN
1  NaN   NaN    NaN
2  NaN   NaN    NaN


can someone please explain the logic or flow of this. I dont really understand why suddenly all values become NaN. dont understand the part when isnull being index to original Df, suddenly the result become NaN all. 

Thank you
2

The syntax for .isnull() is:

df.isnull() 
OR,

df.isnull().sum() 

OR,
 
df.isnull().sum().sum()

Or, on a particular column:

df.col_name.isnull()

Which detects NaN (which are actually floating point values) values in a df and returns a boolean value. .isnull() is the method and you’re calling it on the pandas series, in this case a data frame. As noted above it also works on columns.

What would be the purpose of writing: ‘diabetes_data[diabetes_data.isnull()]’ and pass the data frame object back to the data frame itself?

https://pandas.pydata.org/pandas-docs/stable/reference/api/pandas.DataFrame.isnull.html

3

I am trying to understand the logic of diabetes_data[diabetes_data.isnull()]. like what happened? why can suddenly the value become all NaN

4

.isnull() is the method being called on the df and it then creates a boolean object…and then you’ve used brackets, so it’s being used as an argument on the df and the results are NaN. The brackets [ ] create a series object. The brackets are an index operator as well.

https://pandas.pydata.org/pandas-docs/stable/user_guide/indexing.html