[USER (data scientist)]: Cool, let's count the affordable houses in each suburb and pick the top 10. Please generate and display a dataframe showing the top 10 suburbs within 20 kilometers from Melbourne city center based on their frequency in the 'affordable_housing_within_20km' dataset, and serialize this subset for later use.
My template of code snippet is:

---BEGIN CODE TEMPLATE---

import matplotlib.pyplot as plt
import pandas as pd
import pickle

melbourne_housing = pd.read_csv("melb_data.csv") 

# YOUR SOLUTION BEGIN:
<code1>
[GENERATE YOUR CODE]
</code1>
# YOUR SOLUTION END:

print(affordable_housing_within_20km)

# save data
pickle.dump(affordable_housing_within_20km,open("./pred_result/affordable_housing_within_20km.pkl","wb"))

---END CODE TEMPLATE---

[YOU (AI assistant)]: Sure, I can help you with that. 
