What is the difference between time.clock() and time.time()?

The function time.time() returns the time in seconds since the epoch, i.e., the point where the time starts. For Unix and Windows, the epoch is January 1, 1970 (UTC).

The function time.clock() was used to measure processor time on Unix and wall-clock time on Windows. However, time.clock() was deprecated in Python 3.3 and removed in Python 3.8. The recommended replacements are time.perf_counter() for wall-clock timing and time.process_time() for CPU time.

Syntax

import time

time.time()            # Wall-clock time since epoch
time.perf_counter()    # High-resolution wall-clock timer (replaces time.clock on Windows)
time.process_time()    # CPU time of current process (replaces time.clock on Unix)

time.time()

time.time() returns the wall-clock time in seconds since the epoch. It is used for measuring real elapsed time. Note that time.time() could return a value smaller than a previous call if the system clock has been adjusted between the two calls.

Example

import time

start = time.time()
total = 0
for i in range(1000000):
   total += i
end = time.time()

print("Sum:", total)
print("Wall-clock time:", round(end - start, 6), "seconds")

The output of the above code is −

Sum: 499999500000
Wall-clock time: 0.055498 seconds

time.clock() (Deprecated)

time.clock() had different behavior depending on the platform −

• On Unix − It returned the current processor time (CPU time consumed by the current process), not including time spent sleeping or waiting for I/O.

• On Windows − It returned the wall-clock time elapsed since the first call to time.clock(), based on the Win32 QueryPerformanceCounter() function.

Unlike time.time(), time.clock() always returned non-decreasing values regardless of system clock changes. Due to its inconsistent cross-platform behavior, it was deprecated.

Modern Replacements

Python 3.3 introduced two functions to replace time.clock() with clear, consistent behavior across all platforms −

Example − time.perf_counter()

time.perf_counter() provides a high-resolution wall-clock timer, ideal for benchmarking −

import time

start = time.perf_counter()
total = 0
for i in range(1000000):
   total += i
end = time.perf_counter()

print("Wall-clock time (perf_counter):", round(end - start, 6), "seconds")

The output of the above code is −

Wall-clock time (perf_counter): 0.054321 seconds

Example − time.process_time()

time.process_time() measures only the CPU time used by the current process, excluding sleep and I/O wait −

import time

start = time.process_time()
total = 0
for i in range(1000000):
   total += i
end = time.process_time()

print("CPU time (process_time):", round(end - start, 6), "seconds")

The output of the above code is −

CPU time (process_time): 0.051234 seconds

Comparison

Conclusion

time.time() returns wall-clock time since the epoch and remains widely used. time.clock() was removed due to inconsistent cross-platform behavior. For modern Python, use time.perf_counter() for benchmarking elapsed time and time.process_time() for measuring CPU usage.