Array creation - use Technical Quotient

Converting Python sequences to NumPy Arrays

1. Remember

What does the dtype parameter specify when creating a NumPy array with np.array()?

Dimension of the array
Size of the array
Data type of the array elements

Correct; dtype specifies the data type of the array elements, allowing precise control over the storage format.

Shape of the array

2. Understand

If np.array([127, 128, 129], dtype=np.int8) results in an overflow error, what does this illustrate about NumPy's handling of data types?

NumPy strictly enforces the range limits of specified data types.
NumPy converts all integers to 64-bit by default to avoid overflow.
NumPy automatically adjusts data types to prevent overflow errors.
NumPy ignores data type specifications to optimize memory usage.

Incorrect; NumPy respects dtype specifications for precise control over memory usage and performance.

3. Analyze

You are working on a data analysis project and you decide to perform subtraction between two NumPy arrays, a = np.array([2, 3, 4], dtype=np.uint32) and b = np.array([5, 6, 7], dtype=np.uint32). Surprisingly, the operation a - b yields values like 4294967293. This outcome seems incorrect at first glance. What is the underlying reason for such a result in this calculation?

NumPy arrays cannot perform subtraction on unsigned integers.
The operation incorrectly calculates due to a bug in NumPy.
The subtraction operation converts all integers to floating-point numbers.
Unsigned integer underflow occurs because subtraction results fall below zero.

4. Analyze

You are working with two NumPy arrays for a financial application, a = np.array([2, 3, 4], dtype=np.uint32) represents account credits and b = np.array([5, 6, 7], dtype=np.uint32) represents account debits. When calculating the net movement by executing c_signed32 = a - b.astype(np.int32), resulting in [-3, -3, -3], this operation highlights a crucial NumPy behavior. What does this behavior illustrate about how NumPy handles arithmetic operations involving arrays with different data types?

NumPy defaults to using unsigned integers for all arithmetic operations.
Operations with different data types are not allowed and result in an error.
NumPy always converts the result of arithmetic operations to floating-point numbers.
NumPy promotes the resulting array to a data type that can represent both input arrays' elements.

5. Apply

You need to safely subtract two arrays, x = np.array([1000, 2000, 3000], dtype=np.uint16) and y = np.array([500, 1500, 2500], dtype=np.uint16), to avoid underflow. How would you perform this operation?

Convert both arrays to a signed integer type before subtracting.

Correct; converting both arrays to a signed integer type, such as np.int32, before performing the subtraction avoids underflow by allowing negative results.

Subtract y from x without changing data types.
Increase the size of x and y to prevent underflow.
Use np.subtract(x, y, dtype=np.int32).

Incorrect; np.subtract does not have a dtype parameter to directly specify the result's data type; the operands' types dictate the operation's behavior.

6. Apply

To analyze temperature variations, you have temperature data in Fahrenheit in an array f_temps = np.array([32, 212, 0]). You need to convert this data to Celsius using the formula C = (F - 32) * 5/9. How would you accomplish this with NumPy?

Store the Fahrenheit temperatures in a list, then convert to a NumPy array.

c_temps = (f_temps - 32) * 5 / 9

Convert f_temps to Celsius manually for each element.

c_temps = np.fToC(f_temps)