How to write a code for the lambda method with pandas and random? - python-3.x

Question

So I am working on this python code where I have imported a excel file in from my desktop. Currently in my code I am having problem with executing my lambda function. I am to use the lambda function in the apply() method for the data column(s) with outliers (i.e., z-scores > 3 or < -3), set the value for the outliers to null (i.e., np.nan), otherwise retain the original value.
I have types the line of code 2 different ways with the same outcome. However when I check to see if my overall code used the lambda function it did not and I don't know how to get my code to use the function instead of just running lambda as no part of the overall code.
This is the individual array for each datatypes after using the unique() method.
Have already replace the strings in the array
The image shows the use of the sort_value() method used in my code as well as finding possible outliers for the z-score of the variable before inputting the lambda function into the code.
data.SalaryZScores.apply(lambda SalaryZScores:np.nan if SalaryZScores > 3 or SalaryZScores<-3 else SalaryZScores)
Or
data['Salary'] = data['Salary'].apply(lambda Salary: random.randrange(data['Salary'].min(), data['Salary'].max())if pd.isnull(Salary) else Salary)

Answer 1

It's difficult to understand what your snippets are doing as you haven't provided the code to produce any of the objects in your question. If I understand what you're trying to do the following approaches might solve your problem:
Using .apply() and lambda
import pandas as pd
import numpy as np
# Create some data
data = np.arange(20).reshape(10, 2)
df = pd.DataFrame(data, columns=list('AB'))
df
[Out]
A B
0 0 1
1 2 3
2 4 5
3 6 7
4 8 9
5 10 11
6 12 13
7 14 15
8 16 17
9 18 19
df['A'] = df['A'].apply(lambda x : np.nan if (x < 4) or (x > 10) else x)
df
[Out]
A B
0 NaN 1
1 NaN 3
2 4.0 5
3 6.0 7
4 8.0 9
5 10.0 11
6 NaN 13
7 NaN 15
8 NaN 17
9 NaN 19
Using Series.between:
df['A'].between(4, 10) # Creates a boolean mask between 4 and 10
[Out]
A
0 False
1 False
2 True
3 True
4 True
5 True
6 False
7 False
8 False
9 False
df['A'][~df['A'].between(4, 10)] = np.nan # Uses ~ to invert the boolean mask and set values to np.nan
df
[Out]
A B
0 NaN 1
1 NaN 3
2 4.0 5
3 6.0 7
4 8.0 9
5 10.0 11
6 NaN 13
7 NaN 15
8 NaN 17
9 NaN 19