pyerualjetwork 4.1.4__tar.gz → 4.1.6__tar.gz
This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
- {pyerualjetwork-4.1.4 → pyerualjetwork-4.1.6}/PKG-INFO +35 -10
- {pyerualjetwork-4.1.4 → pyerualjetwork-4.1.6}/README.md +35 -10
- {pyerualjetwork-4.1.4 → pyerualjetwork-4.1.6}/pyerualjetwork/__init__.py +5 -4
- {pyerualjetwork-4.1.4 → pyerualjetwork-4.1.6}/pyerualjetwork/activation_functions_cuda.py +1 -1
- {pyerualjetwork-4.1.4 → pyerualjetwork-4.1.6}/pyerualjetwork/data_operations.py +41 -80
- {pyerualjetwork-4.1.4 → pyerualjetwork-4.1.6}/pyerualjetwork/data_operations_cuda.py +48 -92
- pyerualjetwork-4.1.6/pyerualjetwork/memory_operations.py +303 -0
- {pyerualjetwork-4.1.4 → pyerualjetwork-4.1.6}/pyerualjetwork/metrics_cuda.py +3 -4
- {pyerualjetwork-4.1.4 → pyerualjetwork-4.1.6}/pyerualjetwork/model_operations.py +4 -5
- {pyerualjetwork-4.1.4 → pyerualjetwork-4.1.6}/pyerualjetwork/model_operations_cuda.py +7 -6
- {pyerualjetwork-4.1.4 → pyerualjetwork-4.1.6}/pyerualjetwork/plan.py +35 -23
- {pyerualjetwork-4.1.4 → pyerualjetwork-4.1.6}/pyerualjetwork/plan_cuda.py +113 -81
- {pyerualjetwork-4.1.4 → pyerualjetwork-4.1.6}/pyerualjetwork/visualizations.py +147 -125
- {pyerualjetwork-4.1.4 → pyerualjetwork-4.1.6}/pyerualjetwork/visualizations_cuda.py +193 -169
- {pyerualjetwork-4.1.4 → pyerualjetwork-4.1.6}/pyerualjetwork.egg-info/PKG-INFO +35 -10
- {pyerualjetwork-4.1.4 → pyerualjetwork-4.1.6}/pyerualjetwork.egg-info/SOURCES.txt +1 -0
- {pyerualjetwork-4.1.4 → pyerualjetwork-4.1.6}/setup.py +1 -1
- {pyerualjetwork-4.1.4 → pyerualjetwork-4.1.6}/pyerualjetwork/activation_functions.py +0 -0
- {pyerualjetwork-4.1.4 → pyerualjetwork-4.1.6}/pyerualjetwork/help.py +0 -0
- {pyerualjetwork-4.1.4 → pyerualjetwork-4.1.6}/pyerualjetwork/loss_functions.py +0 -0
- {pyerualjetwork-4.1.4 → pyerualjetwork-4.1.6}/pyerualjetwork/loss_functions_cuda.py +0 -0
- {pyerualjetwork-4.1.4 → pyerualjetwork-4.1.6}/pyerualjetwork/metrics.py +0 -0
- {pyerualjetwork-4.1.4 → pyerualjetwork-4.1.6}/pyerualjetwork/planeat.py +0 -0
- {pyerualjetwork-4.1.4 → pyerualjetwork-4.1.6}/pyerualjetwork/planeat_cuda.py +0 -0
- {pyerualjetwork-4.1.4 → pyerualjetwork-4.1.6}/pyerualjetwork/ui.py +0 -0
- {pyerualjetwork-4.1.4 → pyerualjetwork-4.1.6}/pyerualjetwork.egg-info/dependency_links.txt +0 -0
- {pyerualjetwork-4.1.4 → pyerualjetwork-4.1.6}/pyerualjetwork.egg-info/top_level.txt +0 -0
- {pyerualjetwork-4.1.4 → pyerualjetwork-4.1.6}/setup.cfg +0 -0
|
@@ -1,17 +1,26 @@
|
|
|
1
1
|
Metadata-Version: 2.1
|
|
2
2
|
Name: pyerualjetwork
|
|
3
|
-
Version: 4.1.
|
|
3
|
+
Version: 4.1.6
|
|
4
4
|
Summary: PyerualJetwork is a machine learning library written in Python for professionals, incorporating advanced, unique, new, and modern techniques.
|
|
5
5
|
Author: Hasan Can Beydili
|
|
6
6
|
Author-email: tchasancan@gmail.com
|
|
7
7
|
Keywords: model evaluation,classification,potentiation learning artificial neural networks,NEAT,genetic algorithms,reinforcement learning,neural networks
|
|
8
8
|
Description-Content-Type: text/markdown
|
|
9
9
|
|
|
10
|
-
# PyerualJetwork [](https://socket.dev/pypi/package/pyerualjetwork/overview/4.0.6/tar-gz) [](https://www.codefactor.io/repository/github/hcb06/pyerualjetwork) [](https://socket.dev/pypi/package/pyerualjetwork/overview/4.0.6/tar-gz) [](https://www.codefactor.io/repository/github/hcb06/pyerualjetwork) [](https://pepy.tech/projects/pyerualjetwork) + [](https://pepy.tech/projects/anaplan)
|
|
11
|
+
|
|
12
|
+
|
|
13
|
+
[](https://pepy.tech/projects/pyerualjetwork) [](https://pepy.tech/projects/pyerualjetwork) [](https://pypi.org/project/pyerualjetwork/)
|
|
11
14
|
|
|
12
15
|
Note: anaplan old name of pyerualjetwork
|
|
13
16
|
|
|
14
|
-
https://
|
|
17
|
+
<br><br><br>
|
|
18
|
+
|
|
19
|
+
Libraries.io Page: https://libraries.io/pypi/pyerualjetwork
|
|
20
|
+
|
|
21
|
+
PyPi Page: https://pypi.org/project/pyerualjetwork/
|
|
22
|
+
|
|
23
|
+
GitHub Page: https://github.com/HCB06/PyerualJetwork
|
|
15
24
|
|
|
16
25
|
|
|
17
26
|
pip install pyerualjetwork
|
|
@@ -29,7 +38,7 @@ https://libraries.io/pypi/pyerualjetwork
|
|
|
29
38
|
Optimized for Visual Studio Code
|
|
30
39
|
|
|
31
40
|
requires=[
|
|
32
|
-
'setuptools==75.6.0'
|
|
41
|
+
'setuptools==75.6.0',
|
|
33
42
|
'scipy==1.13.1',
|
|
34
43
|
'tqdm==4.66.4',
|
|
35
44
|
'seaborn==0.13.2',
|
|
@@ -38,27 +47,31 @@ https://libraries.io/pypi/pyerualjetwork
|
|
|
38
47
|
'numpy==1.26.4',
|
|
39
48
|
'matplotlib==3.9.0',
|
|
40
49
|
'colorama==0.4.6',
|
|
41
|
-
'cupy-cuda12x'
|
|
50
|
+
'cupy-cuda12x',
|
|
51
|
+
'psutil==6.1.1'
|
|
42
52
|
]
|
|
43
53
|
|
|
44
54
|
matplotlib, seaborn, networkx (optional).
|
|
45
|
-
PyerualJetwork checks and install all dependencies
|
|
55
|
+
PyerualJetwork checks and install all dependencies with optional ones for every runing.
|
|
46
56
|
If your version is higher or lower, PyerualJetwork automaticly delete other versions and installs this versions.
|
|
47
57
|
|
|
48
58
|
##############################
|
|
49
59
|
|
|
50
60
|
ABOUT PYERUALJETWORK:
|
|
51
61
|
|
|
52
|
-
PyerualJetwork is a machine learning library written in Python for professionals, incorporating advanced, unique, new, and modern techniques. Its most important component is the PLAN (Potentiation Learning Artificial Neural Network) https://papers.ssrn.com/sol3/papers.cfm?abstract_id=4862342. (THIS ARTICLE IS FIRST VERSION OF PLAN.) MODERN VERSION OF PLAN: https://github.com/HCB06/PyerualJetwork/blob/main/Welcome_to_PLAN/PLAN.pdf
|
|
62
|
+
PyerualJetwork is a machine learning library written in Python for professionals, incorporating advanced, unique, new, and modern techniques with optimized GPU acceleration. Its most important component is the PLAN (Potentiation Learning Artificial Neural Network) https://papers.ssrn.com/sol3/papers.cfm?abstract_id=4862342. (THIS ARTICLE IS FIRST VERSION OF PLAN.) MODERN VERSION OF PLAN: https://github.com/HCB06/PyerualJetwork/blob/main/Welcome_to_PLAN/PLAN.pdf
|
|
53
63
|
Both the PLAN algorithm and the PyerualJetwork library were created by Author, and all rights are reserved by Author.
|
|
54
64
|
PyerualJetwork is free to use for commercial business and individual users. PyerualJetwork is written in fully functional programming with non-oop elements. PyerualJetwork consists of many functions that complement each other, which facilitates the learning process and debugging during use.
|
|
55
65
|
As of 12/21/2024, the library includes PLAN and PLANEAT module, but other machine learning modules are expected to be added in the future.
|
|
66
|
+
|
|
67
|
+
PyerualJetwork ready for both eager execution(like PyTorch) and static graph(like Tensorflow) concepts because PyerualJetwork using only functions.
|
|
68
|
+
For example:
|
|
69
|
+
|
|
70
|
+
fit function only fits given training data(suitable for dynamic graph) but learner function learns and optimize entire architecture(suitable for static graph). Or more deeper eager executions PyerualJetwork have: feed_forward function, list of activation functions, loss functions. You can create your unique model architecture. Move your data to GPU or CPU or manage how much should in GPU, Its all up to you.
|
|
56
71
|
<br><br>
|
|
57
72
|
|
|
58
73
|
PyerualJetworket includes Plan Vision, NLPlan, PLANEAT and at the between of both, Deep Plan.<br>
|
|
59
74
|
|
|
60
|
-
<br><br><br>
|
|
61
|
-
|
|
62
75
|
PLAN VISION:<br>
|
|
63
76
|
|
|
64
77
|
|
|
@@ -82,8 +95,20 @@ PLANEAT:<br>
|
|
|
82
95
|
You can create artificial intelligence models that perform reinforcement learning tasks and genetic optimization tasks using the planeat module:
|
|
83
96
|
|
|
84
97
|
<br>
|
|
85
|
-
|
|
86
98
|
<br>
|
|
99
|
+
<br><br>
|
|
100
|
+
|
|
101
|
+
YOU CAN CREATE DYNAMIC ANIMATIONS OF YOUR MODELS
|
|
102
|
+
|
|
103
|
+
<br>
|
|
104
|
+
<br>
|
|
105
|
+
<br>
|
|
106
|
+
|
|
107
|
+
YOU CAN CREATE AND VISUALIZE YOUR MODEL ARCHITECTURE
|
|
108
|
+
|
|
109
|
+
<br>
|
|
110
|
+
<br>
|
|
111
|
+
|
|
87
112
|
|
|
88
113
|
|
|
89
114
|
HOW DO I IMPORT IT TO MY PROJECT?
|
|
@@ -1,8 +1,17 @@
|
|
|
1
|
-
# PyerualJetwork [](https://socket.dev/pypi/package/pyerualjetwork/overview/4.0.6/tar-gz) [](https://www.codefactor.io/repository/github/hcb06/pyerualjetwork) [](https://socket.dev/pypi/package/pyerualjetwork/overview/4.0.6/tar-gz) [](https://www.codefactor.io/repository/github/hcb06/pyerualjetwork) [](https://pepy.tech/projects/pyerualjetwork) + [](https://pepy.tech/projects/anaplan)
|
|
2
|
+
|
|
3
|
+
|
|
4
|
+
[](https://pepy.tech/projects/pyerualjetwork) [](https://pepy.tech/projects/pyerualjetwork) [](https://pypi.org/project/pyerualjetwork/)
|
|
2
5
|
|
|
3
6
|
Note: anaplan old name of pyerualjetwork
|
|
4
7
|
|
|
5
|
-
https://
|
|
8
|
+
<br><br><br>
|
|
9
|
+
|
|
10
|
+
Libraries.io Page: https://libraries.io/pypi/pyerualjetwork
|
|
11
|
+
|
|
12
|
+
PyPi Page: https://pypi.org/project/pyerualjetwork/
|
|
13
|
+
|
|
14
|
+
GitHub Page: https://github.com/HCB06/PyerualJetwork
|
|
6
15
|
|
|
7
16
|
|
|
8
17
|
pip install pyerualjetwork
|
|
@@ -20,7 +29,7 @@ https://libraries.io/pypi/pyerualjetwork
|
|
|
20
29
|
Optimized for Visual Studio Code
|
|
21
30
|
|
|
22
31
|
requires=[
|
|
23
|
-
'setuptools==75.6.0'
|
|
32
|
+
'setuptools==75.6.0',
|
|
24
33
|
'scipy==1.13.1',
|
|
25
34
|
'tqdm==4.66.4',
|
|
26
35
|
'seaborn==0.13.2',
|
|
@@ -29,27 +38,31 @@ https://libraries.io/pypi/pyerualjetwork
|
|
|
29
38
|
'numpy==1.26.4',
|
|
30
39
|
'matplotlib==3.9.0',
|
|
31
40
|
'colorama==0.4.6',
|
|
32
|
-
'cupy-cuda12x'
|
|
41
|
+
'cupy-cuda12x',
|
|
42
|
+
'psutil==6.1.1'
|
|
33
43
|
]
|
|
34
44
|
|
|
35
45
|
matplotlib, seaborn, networkx (optional).
|
|
36
|
-
PyerualJetwork checks and install all dependencies
|
|
46
|
+
PyerualJetwork checks and install all dependencies with optional ones for every runing.
|
|
37
47
|
If your version is higher or lower, PyerualJetwork automaticly delete other versions and installs this versions.
|
|
38
48
|
|
|
39
49
|
##############################
|
|
40
50
|
|
|
41
51
|
ABOUT PYERUALJETWORK:
|
|
42
52
|
|
|
43
|
-
PyerualJetwork is a machine learning library written in Python for professionals, incorporating advanced, unique, new, and modern techniques. Its most important component is the PLAN (Potentiation Learning Artificial Neural Network) https://papers.ssrn.com/sol3/papers.cfm?abstract_id=4862342. (THIS ARTICLE IS FIRST VERSION OF PLAN.) MODERN VERSION OF PLAN: https://github.com/HCB06/PyerualJetwork/blob/main/Welcome_to_PLAN/PLAN.pdf
|
|
53
|
+
PyerualJetwork is a machine learning library written in Python for professionals, incorporating advanced, unique, new, and modern techniques with optimized GPU acceleration. Its most important component is the PLAN (Potentiation Learning Artificial Neural Network) https://papers.ssrn.com/sol3/papers.cfm?abstract_id=4862342. (THIS ARTICLE IS FIRST VERSION OF PLAN.) MODERN VERSION OF PLAN: https://github.com/HCB06/PyerualJetwork/blob/main/Welcome_to_PLAN/PLAN.pdf
|
|
44
54
|
Both the PLAN algorithm and the PyerualJetwork library were created by Author, and all rights are reserved by Author.
|
|
45
55
|
PyerualJetwork is free to use for commercial business and individual users. PyerualJetwork is written in fully functional programming with non-oop elements. PyerualJetwork consists of many functions that complement each other, which facilitates the learning process and debugging during use.
|
|
46
56
|
As of 12/21/2024, the library includes PLAN and PLANEAT module, but other machine learning modules are expected to be added in the future.
|
|
57
|
+
|
|
58
|
+
PyerualJetwork ready for both eager execution(like PyTorch) and static graph(like Tensorflow) concepts because PyerualJetwork using only functions.
|
|
59
|
+
For example:
|
|
60
|
+
|
|
61
|
+
fit function only fits given training data(suitable for dynamic graph) but learner function learns and optimize entire architecture(suitable for static graph). Or more deeper eager executions PyerualJetwork have: feed_forward function, list of activation functions, loss functions. You can create your unique model architecture. Move your data to GPU or CPU or manage how much should in GPU, Its all up to you.
|
|
47
62
|
<br><br>
|
|
48
63
|
|
|
49
64
|
PyerualJetworket includes Plan Vision, NLPlan, PLANEAT and at the between of both, Deep Plan.<br>
|
|
50
65
|
|
|
51
|
-
<br><br><br>
|
|
52
|
-
|
|
53
66
|
PLAN VISION:<br>
|
|
54
67
|
|
|
55
68
|
|
|
@@ -73,8 +86,20 @@ PLANEAT:<br>
|
|
|
73
86
|
You can create artificial intelligence models that perform reinforcement learning tasks and genetic optimization tasks using the planeat module:
|
|
74
87
|
|
|
75
88
|
<br>
|
|
76
|
-
|
|
77
89
|
<br>
|
|
90
|
+
<br><br>
|
|
91
|
+
|
|
92
|
+
YOU CAN CREATE DYNAMIC ANIMATIONS OF YOUR MODELS
|
|
93
|
+
|
|
94
|
+
<br>
|
|
95
|
+
<br>
|
|
96
|
+
<br>
|
|
97
|
+
|
|
98
|
+
YOU CAN CREATE AND VISUALIZE YOUR MODEL ARCHITECTURE
|
|
99
|
+
|
|
100
|
+
<br>
|
|
101
|
+
<br>
|
|
102
|
+
|
|
78
103
|
|
|
79
104
|
|
|
80
105
|
HOW DO I IMPORT IT TO MY PROJECT?
|
|
@@ -83,4 +108,4 @@ Anaconda users can access the 'Anaconda Prompt' terminal from the Start menu and
|
|
|
83
108
|
|
|
84
109
|
After installing the module using "pip" you can now call the library module in your project environment. Use: “from pyerualjetwork import plan”. Now, you can call the necessary functions from the plan module.
|
|
85
110
|
|
|
86
|
-
The PLAN algorithm will not be explained in this document. This document focuses on how professionals can integrate and use PyerualJetwork in their systems. However, briefly, the PLAN algorithm can be described as a classification algorithm. PLAN algorithm achieves this task with an incredibly energy-efficient, fast, and hyperparameter-free user-friendly approach. For more detailed information, you can check out .pdf) file.
|
|
111
|
+
The PLAN algorithm will not be explained in this document. This document focuses on how professionals can integrate and use PyerualJetwork in their systems. However, briefly, the PLAN algorithm can be described as a classification algorithm. PLAN algorithm achieves this task with an incredibly energy-efficient, fast, and hyperparameter-free user-friendly approach. For more detailed information, you can check out .pdf) file.
|
|
@@ -2,7 +2,6 @@
|
|
|
2
2
|
import subprocess
|
|
3
3
|
subprocess.check_call(["pip", "install", 'setuptools==75.6.0'])
|
|
4
4
|
import pkg_resources
|
|
5
|
-
from datetime import datetime
|
|
6
5
|
|
|
7
6
|
print("Auto checking and installation dependencies for PyerualJetwork")
|
|
8
7
|
|
|
@@ -14,7 +13,9 @@ package_names = [
|
|
|
14
13
|
'networkx==3.3',
|
|
15
14
|
'numpy==1.26.4',
|
|
16
15
|
'matplotlib==3.9.0',
|
|
17
|
-
'colorama==0.4.6'
|
|
16
|
+
'colorama==0.4.6',
|
|
17
|
+
'psutil==6.1.1',
|
|
18
|
+
'cupy-cuda12x==13.3.0'
|
|
18
19
|
]
|
|
19
20
|
|
|
20
21
|
installed_packages = pkg_resources.working_set
|
|
@@ -47,8 +48,8 @@ for package_name in package_names:
|
|
|
47
48
|
|
|
48
49
|
print(f"PyerualJetwork is ready to use with {err} errors")
|
|
49
50
|
|
|
50
|
-
__version__ = "4.1.
|
|
51
|
-
__update__ = "*
|
|
51
|
+
__version__ = "4.1.6"
|
|
52
|
+
__update__ = "* Changes: https://github.com/HCB06/PyerualJetwork/blob/main/CHANGES\n* PyerualJetwork document: https://github.com/HCB06/PyerualJetwork/blob/main/Welcome_to_PyerualJetwork/PYERUALJETWORK_USER_MANUEL_AND_LEGAL_INFORMATION(EN).pdf\n* YouTube tutorials: https://www.youtube.com/@HasanCanBeydili"
|
|
52
53
|
|
|
53
54
|
def print_version(__version__):
|
|
54
55
|
print(f"PyerualJetwork Version {__version__}" + '\n')
|
|
@@ -16,26 +16,10 @@ def encode_one_hot(y_train, y_test=None, summary=False):
|
|
|
16
16
|
Returns:
|
|
17
17
|
tuple: One-hot encoded y_train and (if given) y_test.
|
|
18
18
|
"""
|
|
19
|
-
|
|
20
|
-
|
|
21
|
-
|
|
22
|
-
|
|
23
|
-
if y_train.dtype != np.uint16:
|
|
24
|
-
y_train = np.array(y_train, copy=False).astype(np.uint16, copy=False)
|
|
25
|
-
else:
|
|
26
|
-
if y_train.dtype != np.uint32:
|
|
27
|
-
y_train = np.array(y_train, copy=False).astype(np.uint32, copy=False)
|
|
28
|
-
|
|
29
|
-
if y_test is not None:
|
|
30
|
-
if len(y_test) < 256:
|
|
31
|
-
if y_test.dtype != np.uint8:
|
|
32
|
-
y_test = np.array(y_test, copy=False).astype(np.uint8, copy=False)
|
|
33
|
-
elif len(y_test) <= 32767:
|
|
34
|
-
if y_test.dtype != np.uint16:
|
|
35
|
-
y_test = np.array(y_test, copy=False).astype(np.uint16, copy=False)
|
|
36
|
-
else:
|
|
37
|
-
if y_test.dtype != np.uint32:
|
|
38
|
-
y_test = np.array(y_test, copy=False).astype(np.uint32, copy=False)
|
|
19
|
+
from .memory_operations import optimize_labels
|
|
20
|
+
|
|
21
|
+
y_train = optimize_labels(y_train, one_hot_encoded=False, cuda=False)
|
|
22
|
+
y_test = optimize_labels(y_test, one_hot_encoded=False, cuda=False)
|
|
39
23
|
|
|
40
24
|
classes = np.unique(y_train)
|
|
41
25
|
class_count = len(classes)
|
|
@@ -47,12 +31,12 @@ def encode_one_hot(y_train, y_test=None, summary=False):
|
|
|
47
31
|
for cls, idx in class_to_index.items():
|
|
48
32
|
print(f" {idx}: {cls}")
|
|
49
33
|
|
|
50
|
-
y_train_encoded = np.zeros((y_train.shape[0], class_count))
|
|
34
|
+
y_train_encoded = np.zeros((y_train.shape[0], class_count), dtype=y_train.dtype)
|
|
51
35
|
for i, label in enumerate(y_train):
|
|
52
36
|
y_train_encoded[i, class_to_index[label]] = 1
|
|
53
37
|
|
|
54
38
|
if y_test is not None:
|
|
55
|
-
y_test_encoded = np.zeros((y_test.shape[0], class_count))
|
|
39
|
+
y_test_encoded = np.zeros((y_test.shape[0], class_count), dtype=y_test.dtype)
|
|
56
40
|
for i, label in enumerate(y_test):
|
|
57
41
|
y_test_encoded[i, class_to_index[label]] = 1
|
|
58
42
|
return y_train_encoded, y_test_encoded
|
|
@@ -90,16 +74,10 @@ def split(X, y, test_size, random_state=42, dtype=np.float32):
|
|
|
90
74
|
Returns:
|
|
91
75
|
tuple: x_train, x_test, y_train, y_test as ordered training and testing data subsets.
|
|
92
76
|
"""
|
|
93
|
-
|
|
94
|
-
|
|
95
|
-
|
|
96
|
-
|
|
97
|
-
elif len(y) <= 32767:
|
|
98
|
-
if y.dtype != np.uint16:
|
|
99
|
-
y = np.array(y, copy=False).astype(np.uint16, copy=False)
|
|
100
|
-
else:
|
|
101
|
-
if y.dtype != np.uint32:
|
|
102
|
-
y = np.array(y, copy=False).astype(np.uint32, copy=False)
|
|
77
|
+
from .memory_operations import transfer_to_cpu, optimize_labels
|
|
78
|
+
|
|
79
|
+
X = transfer_to_cpu(X, dtype=dtype)
|
|
80
|
+
y = optimize_labels(y, one_hot_encoded=False, cuda=False)
|
|
103
81
|
|
|
104
82
|
num_samples = X.shape[0]
|
|
105
83
|
|
|
@@ -124,6 +102,8 @@ def split(X, y, test_size, random_state=42, dtype=np.float32):
|
|
|
124
102
|
x_train, x_test = X[train_indices], X[test_indices]
|
|
125
103
|
y_train, y_test = y[train_indices], y[test_indices]
|
|
126
104
|
|
|
105
|
+
del X, y
|
|
106
|
+
|
|
127
107
|
return x_train, x_test, y_train, y_test
|
|
128
108
|
|
|
129
109
|
|
|
@@ -142,20 +122,11 @@ def manuel_balancer(x_train, y_train, target_samples_per_class, dtype=np.float32
|
|
|
142
122
|
y_balanced -- Balanced class labels (one-hot encoded, NumPy array format)
|
|
143
123
|
"""
|
|
144
124
|
from .ui import loading_bars
|
|
145
|
-
|
|
146
|
-
bar_format = loading_bars()[0]
|
|
125
|
+
from .memory_operations import transfer_to_cpu
|
|
147
126
|
|
|
148
|
-
x_train =
|
|
149
|
-
if len(y_train[0]) < 256:
|
|
150
|
-
if y_train.dtype != np.uint8:
|
|
151
|
-
y_train = np.array(y_train, copy=False).astype(np.uint8, copy=False)
|
|
152
|
-
elif len(y_train[0]) <= 32767:
|
|
153
|
-
if y_train.dtype != np.uint16:
|
|
154
|
-
y_train = np.array(y_train, copy=False).astype(np.uint16, copy=False)
|
|
155
|
-
else:
|
|
156
|
-
if y_train.dtype != np.uint32:
|
|
157
|
-
y_train = np.array(y_train, copy=False).astype(np.uint32, copy=False)
|
|
127
|
+
x_train = transfer_to_cpu(x_train, dtype=dtype)
|
|
158
128
|
|
|
129
|
+
bar_format = loading_bars()[0]
|
|
159
130
|
classes = np.arange(y_train.shape[1])
|
|
160
131
|
class_count = len(classes)
|
|
161
132
|
|
|
@@ -181,8 +152,8 @@ def manuel_balancer(x_train, y_train, target_samples_per_class, dtype=np.float32
|
|
|
181
152
|
if num_samples < target_samples_per_class:
|
|
182
153
|
|
|
183
154
|
samples_to_add = target_samples_per_class - num_samples
|
|
184
|
-
additional_samples = np.zeros((samples_to_add, x_train.shape[1]))
|
|
185
|
-
additional_labels = np.zeros((samples_to_add, y_train.shape[1]))
|
|
155
|
+
additional_samples = np.zeros((samples_to_add, x_train.shape[1]), dtype=x_train.dtype)
|
|
156
|
+
additional_labels = np.zeros((samples_to_add, y_train.shape[1]), dtype=y_train.dtype)
|
|
186
157
|
|
|
187
158
|
for i in range(samples_to_add):
|
|
188
159
|
|
|
@@ -200,10 +171,12 @@ def manuel_balancer(x_train, y_train, target_samples_per_class, dtype=np.float32
|
|
|
200
171
|
x_balanced.append(additional_samples)
|
|
201
172
|
y_balanced.append(additional_labels)
|
|
202
173
|
|
|
203
|
-
x_balanced = np.vstack(x_balanced)
|
|
204
|
-
y_balanced = np.vstack(y_balanced)
|
|
174
|
+
x_balanced = np.vstack(x_balanced, dtype=x_train.dtype)
|
|
175
|
+
y_balanced = np.vstack(y_balanced, dtype=y_train.dtype)
|
|
176
|
+
|
|
177
|
+
del x_train, y_train
|
|
205
178
|
|
|
206
|
-
return x_balanced
|
|
179
|
+
return x_balanced, y_balanced
|
|
207
180
|
|
|
208
181
|
|
|
209
182
|
def auto_balancer(x_train, y_train, dtype=np.float32):
|
|
@@ -220,20 +193,11 @@ def auto_balancer(x_train, y_train, dtype=np.float32):
|
|
|
220
193
|
tuple: A tuple containing balanced input data and labels.
|
|
221
194
|
"""
|
|
222
195
|
from .ui import loading_bars
|
|
223
|
-
|
|
224
|
-
bar_format = loading_bars()[0]
|
|
196
|
+
from .memory_operations import transfer_to_cpu
|
|
225
197
|
|
|
226
|
-
x_train =
|
|
227
|
-
if len(y_train[0]) < 256:
|
|
228
|
-
if y_train.dtype != np.uint8:
|
|
229
|
-
y_train = np.array(y_train, copy=False).astype(np.uint8, copy=False)
|
|
230
|
-
elif len(y_train[0]) <= 32767:
|
|
231
|
-
if y_train.dtype != np.uint16:
|
|
232
|
-
y_train = np.array(y_train, copy=False).astype(np.uint16, copy=False)
|
|
233
|
-
else:
|
|
234
|
-
if y_train.dtype != np.uint32:
|
|
235
|
-
y_train = np.array(y_train, copy=False).astype(np.uint32, copy=False)
|
|
198
|
+
x_train = transfer_to_cpu(x_train, dtype=dtype)
|
|
236
199
|
|
|
200
|
+
bar_format = loading_bars()[0]
|
|
237
201
|
classes = np.arange(y_train.shape[1])
|
|
238
202
|
class_count = len(classes)
|
|
239
203
|
|
|
@@ -271,10 +235,15 @@ def auto_balancer(x_train, y_train, dtype=np.float32):
|
|
|
271
235
|
print(Fore.RED + "ERROR: Inputs and labels must be same length check parameters")
|
|
272
236
|
sys.exit()
|
|
273
237
|
|
|
274
|
-
|
|
238
|
+
BalancedInputs = BalancedInputs.astype(dtype, copy=False)
|
|
239
|
+
BalancedLabels = BalancedLabels.astype(dtype=y_train.dtype, copy=False)
|
|
275
240
|
|
|
241
|
+
del x_train, y_train
|
|
276
242
|
|
|
277
|
-
|
|
243
|
+
return BalancedInputs, BalancedLabels
|
|
244
|
+
|
|
245
|
+
|
|
246
|
+
def synthetic_augmentation(x, y, dtype=np.float32):
|
|
278
247
|
"""
|
|
279
248
|
Generates synthetic examples to balance classes with fewer examples.
|
|
280
249
|
|
|
@@ -288,24 +257,12 @@ def synthetic_augmentation(x_train, y_train, dtype=np.float32):
|
|
|
288
257
|
y_balanced -- Balanced class labels (one-hot encoded, array format)
|
|
289
258
|
"""
|
|
290
259
|
from .ui import loading_bars
|
|
291
|
-
|
|
292
|
-
bar_format = loading_bars()[0]
|
|
260
|
+
from .memory_operations import transfer_to_cpu
|
|
293
261
|
|
|
294
|
-
|
|
295
|
-
if len(y_train[0]) < 256:
|
|
296
|
-
if y_train.dtype != np.uint8:
|
|
297
|
-
y_train = np.array(y_train, copy=False).astype(np.uint8, copy=False)
|
|
298
|
-
elif len(y_train[0]) <= 32767:
|
|
299
|
-
if y_train.dtype != np.uint16:
|
|
300
|
-
y_train = np.array(y_train, copy=False).astype(np.uint16, copy=False)
|
|
301
|
-
else:
|
|
302
|
-
if y_train.dtype != np.uint32:
|
|
303
|
-
y_train = np.array(y_train, copy=False).astype(np.uint32, copy=False)
|
|
304
|
-
|
|
305
|
-
x = x_train
|
|
306
|
-
y = y_train
|
|
262
|
+
x = transfer_to_cpu(x, dtype=dtype)
|
|
307
263
|
|
|
308
|
-
|
|
264
|
+
bar_format = loading_bars()[0]
|
|
265
|
+
classes = np.arange(y.shape[1])
|
|
309
266
|
class_count = len(classes)
|
|
310
267
|
|
|
311
268
|
class_distribution = {i: 0 for i in range(class_count)}
|
|
@@ -340,8 +297,12 @@ def synthetic_augmentation(x_train, y_train, dtype=np.float32):
|
|
|
340
297
|
|
|
341
298
|
num_samples += 1
|
|
342
299
|
|
|
300
|
+
x_balanced = np.array(x_balanced).astype(dtype, copy=False)
|
|
301
|
+
y_balanced = np.array(y_balanced).astype(dtype=y.dtype, copy=False)
|
|
302
|
+
|
|
303
|
+
del x, y
|
|
343
304
|
|
|
344
|
-
return
|
|
305
|
+
return x_balanced, y_balanced
|
|
345
306
|
|
|
346
307
|
|
|
347
308
|
def standard_scaler(x_train=None, x_test=None, scaler_params=None, dtype=np.float32):
|