array([2.1, 3.6, 4.5, 5.8]) and want to round it to array([2, 4, 4, 6]).Method 1: Using numpy.round() Function
The numpy.round() function is a versatile tool for rounding floats to the nearest integer within a NumPy array. It returns an array with each element rounded to the specified number of decimals. When decimals are set to 0, it’s effectively rounding to the nearest integer.Here’s an example:
import numpy as np array_floats = np.array([2.1, 3.6, 4.5, 5.8]) array_rounded = np.round(array_floats) print(array_rounded)
Output:
[2. 3. 4. 6.] This code creates a NumPy array of floats and then rounds them to the nearest integers. The output is an array of floats that appear as integers; however, to convert them to actual integer data type, you would need an additional operation.Method 2: Using numpy.rint() Function
The numpy.rint() function rounds elements to the nearest integer conservatively, without adjusting the half-way cases up as round() might do. This method is another way to round to the nearest integer numerically.Here’s an example:
import numpy as np array_floats = np.array([2.1, 3.6, 4.5, 5.8]) array_rinted = np.rint(array_floats) print(array_rinted)
Output:
[2. 4. 4. 6.] This code snippet illustrates using the np.rint() function to round the elements of a NumPy array to the nearest integers. The array values are effectively the same as with np.round() but may differ for “half-way” cases.Method 3: Using numpy.floor() and numpy.ceil()
The numpy.floor() function returns the largest integer value less than or equal to each element in an array. The numpy.ceil() function does the opposite, providing the smallest integer greater than or equal to each array element. These can be used effectively when we want to round down or up specifically.Here’s an example:
import numpy as np
array_floats = np.array([2.1, 3.6, 4.5, 5.8])
array_floored = np.floor(array_floats)
array_ceiled = np.ceil(array_floats)
print("Floored:", array_floored)
print("Ceiled:", array_ceiled)Output:
Floored: [2. 3. 4. 5.] Ceiled: [3. 4. 5. 6.] By using np.floor() and np.ceil(), the code snippet demonstrates rounding down and up, respectively. Depending on the need, either of these methods can be used to acquire the largest integer less than the element or the smallest integer greater than the element.Method 4: Using numpy.trunc() Function
The numpy.trunc() function returns the truncated value of the input, basically wiping out the decimal part and returning the integer part of each element in the array. This is somewhat similar to floor for positive numbers and ceil for negative numbers.Here’s an example:
import numpy as np array_floats = np.array([2.1, 3.6, -4.5, -5.8]) array_truncated = np.trunc(array_floats) print(array_truncated)
Output:
[ 2. 3. -4. -5.] This example illustrates how np.trunc() is used to discard the fractional part of the numbers—a form of rounding towards zero, which is especially useful in certain mathematical computations or integer conversions while retaining the sign.Bonus One-Liner Method 5: Using Astype with int
A Pythonic and straightforward approach to converting a NumPy array of floats to integers is by using the astype() method. This function enables you to cast an entire array to a specified type, which, when set to int, would round down each element.Here’s an example:
import numpy as np array_floats = np.array([2.1, 3.6, 4.5, 5.8]) array_as_int = array_floats.astype(int) print(array_as_int)
Output:
[2 3 4 5] With this code snippet, we’ve used the .astype() method to economically convert each float in the array to its corresponding integer representation by truncation, which is a swift and memory-efficient operation for large arrays.Summary/Discussion
- Method 1:
numpy.round(). Good for general rounding to the nearest integer with the option to specify decimal precision. May need additional conversion to an integer type. - Method 2:
numpy.rint(). Very similar to round but may differ in handling half-way cases. Returns an array of floats that also may need type conversion. - Method 3:
numpy.floor()/numpy.ceil(). Tailored for cases when one needs to round down or up specifically. Not suitable for rounding to the nearest integer. - Method 4:
numpy.trunc(). Useful for removing decimal parts, effectively rounding towards zero. Particularly useful with mixed sign arrays. - Method 5:
.astype(int). The most direct method for truncating floats to integers without rounding at all.