Project A1 Hidden Tests

Question 2¶

Q2a¶

• Ensure the final answer has a length of 4

• The third element in the final answer should be 10

Question 4¶

Q4a¶

• Make sure the values in training_data haven’t been altered

• For example, check that the values in "Building Square Feet" haven’t been changed

• If we remove outliers for "Log Sale Price" with a lower bound of 8 and higher bound of 10, the sum of the resulting "Building Square Feet" should be close to 5,791,106

Q4b¶

Hint: consider the x-axis and that most of the data is distributed to the left from $0.5 \times 10^6 = \$500,000$.

Q4c¶

After removing outliers, training_data should have 153,776 rows remaining.

Question 5¶

Q5a¶

A good representative of "Description" column value in training_data is shown below:

“This property, sold on 06/10/2016, is a one-story household located at 104 SAUK TRL. It has a total of 5 rooms, 2 of which are bedrooms, and 1.0 of which are bathrooms.”

Consider what features of houses are specified here.

Q5b¶

• The values of the "Bathrooms" column should sum up to 288,922

• training_data should contain 167,797 entries with fewer than 5 bathrooms

Question 6¶

Q6a¶

There are many ways to approach this question. Using .unique() pandas function is one of the ways to solve it.

Q6b¶

• Make sure your final answer only contains 10 unique neighborhood codes

• The sum of the "Neighborhood Code"(s) should equal 3,824,060

Q6c¶

The three most expensive neighborhood codes (using the median) are neighborhood 44, 94, and 93.

Q6d¶

Using the three most expensive neighborhood codes from Q6c, there should be 1,290 rows belonging to those neighborhoods.

Question 7¶

Q7a¶

training_data should have 4 unique wall materials with value_counts() = [70303, 59125, 35717, 3786]

Q7b¶

The sum of the one-hot encoded columns should equal the number of rows in training_data.