# Performance Measurements

The purpose of this page is to have a link to each student exercise's page. When each student finishes the exercise, creates a link on the table below. In the case of multiple solutions, additional measures may be added to measure the times individually.

# Lists Initialization

Code to initialize the list of names and times.

Testing/Parameters variables may be initialized to standardize the inputs.

``` 1 import numpy as np
2 import numpy.random as rp
3
7 #                  [1, 0,-1],
8 #                  [2, 0,-2],
9 #                  [1, 0,-1]
10 #                           ], dtype=np.float32)
11
12 img = np.random.random_integers(0, 255, (1000, 1000)).astype(np.uint8)
```

# Code entries to measure performance

Each entry must contain an individual block to avoid timeouts. The funcTimer function to measure the time is imported on the code above, and the function to be measured should be passed as parameter. If any error happens, the return is 0, and the result of the function is None.

Entries:

ia636:

Pedro:

Fernando:

Greice:

Lucas:

Wendell:

Luiz-1 - Convolução não Periódica:

Luiz-2:

Giovani:

André Körbes:

Victor texture:

Victor shared:

# Sorted Execution Times

Sorts and shows the name and execution times.

Results:

Student Mpixels Kernel Time Speedup OK
Giovani 1.00 35 3.387 ms 129.8 False
Victor shared 1.00 35 4.565 ms 96.3 False
Fernando Otimizado 1.00 35 5.536 ms 79.4 True
Luiz-1 1.00 35 6.171 ms 71.3 False
Pedro 1.00 35 6.641 ms 66.2 False
Luiz-2 1.00 35 6.758 ms 65.1 True
Wendell 1.00 35 6.955 ms 63.2 False
Victor texture 1.00 35 9.995 ms 44.0 False
Greice 1.00 35 10.962 ms 40.1 False
Fernando Shared 1.00 35 10.969 ms 40.1 True
Adriano 1.00 35 11.735 ms 37.5 True
Lucas 1.00 35 12.102 ms 36.3 True
Andre K 1.00 35 36.714 ms 12.0 False
iapconv Python sequential REFERENCE 1.00 35 439.741 ms 1.0 True
Fernando Direto 0.00 35 inf ms 0.0 False