Exercise

# Testing on circular data

Another special case where it is easy to guess the value of \(r^2\) is when the model does not fit the testing dataset at all. In this case, \(r^2\) takes its lowest possible value `0.0`

.

The plot shows such a testing dataset and model. The testing dataset consists of data arranged in a circle of radius `1.0`

. The `x`

and `y`

co-ordinates of the data is shown on the plot. The model corresponds to a straight line `y=0`

.

As one can easily see, the straight line does not fit the data at all. In this particular case, the value of \(r^2\) is known to be `0.0`

.

Your job is to write a test `test_on_circular_data()`

for the function `model_test()`

that performs this sanity check. `pytest`

, `numpy as np`

, `model_test`

, `sin`

, `cos`

and `pi`

have been imported for you.

Instructions 1/4

**undefined XP**

- Assign
`test_argument`

to a \(8 \times 2\) NumPy array holding the circular testing data shown in the plot, starting with`(1.0, 0.0)`

and moving anticlockwise.