%global _empty_manifest_terminate_build 0
Name:		python-augmentation-engine
Version:	0.1.2
Release:	1
Summary:	Solar Filaments data augmentation demo package
License:	MIT License
URL:		https://bitbucket.org/gsudmlab/augmentation_engine/src/master/
Source0:	https://mirrors.nju.edu.cn/pypi/web/packages/e3/6e/7349c0218e5af7f68c817c6391d161b2295d7f2ba5ad9d063e6c887b0dfc/augmentation_engine-0.1.2.tar.gz
BuildArch:	noarch

Requires:	python3-sortedcontainers
Requires:	python3-opencv-python
Requires:	python3-torchvision
Requires:	python3-pillow
Requires:	python3-pycocotools

%description
```python
pip install augmentation_engine
```
    Requirement already satisfied: augmentation_engine in d:\gsu_assignments\semester_2\dl\augmentation_engine (0.0.1)
    Requirement already satisfied: sortedcontainers in c:\users\shreejaa talla\pycharmprojects\bbso_fa\venv\lib\site-packages (from augmentation_engine) (2.4.0)
    Requirement already satisfied: opencv_python in c:\users\shreejaa talla\pycharmprojects\bbso_fa\venv\lib\site-packages (from augmentation_engine) (4.5.3.56)
    Requirement already satisfied: torchvision in c:\users\shreejaa talla\pycharmprojects\bbso_fa\venv\lib\site-packages (from augmentation_engine) (0.10.0)
    Requirement already satisfied: pillow in c:\users\shreejaa talla\pycharmprojects\bbso_fa\venv\lib\site-packages (from augmentation_engine) (8.3.2)
    Requirement already satisfied: numpy>=1.17.3 in c:\users\shreejaa talla\pycharmprojects\bbso_fa\venv\lib\site-packages (from opencv_python->augmentation_engine) (1.21.2)
    Requirement already satisfied: torch==1.9.0 in c:\users\shreejaa talla\pycharmprojects\bbso_fa\venv\lib\site-packages (from torchvision->augmentation_engine) (1.9.0)
    Requirement already satisfied: typing-extensions in c:\users\shreejaa talla\pycharmprojects\bbso_fa\venv\lib\site-packages (from torch==1.9.0->torchvision->augmentation_engine) (3.10.0.2)
    Note: you may need to restart the kernel to use updated packages.
### Import Required Libraries 
```python
import os
from torchvision import transforms
import matplotlib.pyplot as plt
from filament_augmentation.loader.filament_dataloader import FilamentDataLoader
from filament_augmentation.generator.filament_dataset import FilamentDataset
from filament_augmentation.metadata.filament_metadata import FilamentMetadata
```
**To find out the number of left, right and unidentified chiralities for an interval of time.**
- The code snippet below gives the *chirality distribution*, i.e., the distribution of left, right and unidentified chiralities for an interval of time from "2015-08-01 17:36:15" to "2015-08-09 18:15:17".
- Here the petdata has big bear space observatory(BBSO) full disk solar images from (01-09) aug 2015.
- The format for start and end time should be YYYY-MM-DD HH:MM:SS.
- The ann_file or annotation file is a H-alpha based manually labelled filaments in a json file comes with petdata.
```python
__file__ = 'augmentation_process.ipynb'
bbso_json = os.path.abspath(
        os.path.join(os.path.dirname(__file__), 'petdata', 'bbso_json_data','2015_chir_data.json'))
filamentInfo = FilamentMetadata(ann_file = bbso_json, start_time = '2015-08-01 00:00:15',
                                    end_time = '2015-08-30 23:59:59')
filamentInfo.get_chirality_distribution()
```
    (22, 30, 186)
- In order to generate extra filaments for left, right or unidentified chirality by either balancing the data or getting them in required ratios to train them using an ML algorithm. A filament dataset class should be initialized which is quite similar to that of pytorch dataset class.
- The transform list should be list of torchvision [transformations](https://pytorch.org/vision/0.8/transforms.html) 
- Filament ratio is tuple variable that takes (L,R,U).
### Initializing Filament dataset 
To initialize filament dataset class follow parameters are required:
- bbso_path - BBSO full disk H-alpha solar images comes with petdata, path of the folder.
- ann_file - a H-alpha based manually labelled filaments in a json file comes with petdata.
- The format for start and end time should be YYYY-MM-DD HH:MM:SS.
```python
bbso_path = os.path.abspath(os.path.join(os.path.dirname(__file__), 'petdata', '2015'))
dataset = FilamentDataset(bbso_path = bbso_path, ann_file = bbso_json, 
                          start_time = "2015-08-01 17:36:15", end_time = "2015-08-09 17:36:15")
```
    loading annotations into memory...
    Done (t=0.07s)
    creating index...
    index created!
### Setup transformations for data augmentation
The transformations function can be refered from [torchvision transforms](https://pytorch.org/vision/0.8/transforms.html)
- Here transforms variable should have list of torchvision transforms functions as shown below: 
```python
transforms1 = [
    transforms.ColorJitter(brightness=(0.25,1.25), contrast=(0.25,2.00), saturation=(0.25,2.25)),
    transforms.RandomRotation(15,expand=False,fill=110)
]
```
### Initializing data loader
- dataset = object of filament dataset class.
- batch_size = number of filaments to be generated per batch.
- filament_ratio = tuple of three values, i.e., ratios of left, right and unidentified chirality to be generated in each batch.
- n_batchs = number of batchs.
- transforms = list of torchvision transformations functions
- image_dim = image dimensions if image dimension is -1 then image will not be resize, i.e., output is original image size.
```python
data_loader = FilamentDataLoader(dataset = dataset,batch_size = 3 , filament_ratio = (1, 1, 1),n_batchs = 10, 
                                 transforms = transforms1, image_dim = 224)
```
#### How to generate 3 filament images for every batch with ratio of left as 1, right chirality as 1 and unidentified as 1 for 10 batches with original image dimension and display the images?
```python
data_loader = FilamentDataLoader(dataset = dataset,batch_size = 3 , filament_ratio = (1, 1, 1),
                                 n_batchs = 10, transforms = transforms1, image_dim = -1)
```
#### Batch -1 augmented filament images and their following labels (1=R, 0= U, -1=L)
```python
for original_imgs, transformed_imgs, labels in data_loader:
    for org_img, img, label in zip(original_imgs ,transformed_imgs, labels):
        print("Original image")
        plt.imshow(org_img, cmap='gray')
        plt.show()
        print("Transformed image")
        plt.imshow(img, cmap='gray')
        plt.show()
        print("Label",label)
    break
```
    Original image
![png](document_images/output_18_1.png)
    Transformed image
![png](document_images/output_18_3.png)
    Label 0
    Original image
![png](document_images/output_18_5.png)
    Transformed image
![png](document_images/output_18_7.png)
    Label 1
    Original image
![png](document_images/output_18_9.png)
    Transformed image
![png](document_images/output_18_11.png)
    Label -1
#### How to generate 12  filament images for every batch with ratio of left as 2, right chirality as 3  and unidentified as 1 for 5 batches with image dimension of 224x224 ?
```python
data_loader = FilamentDataLoader(dataset = dataset,batch_size = 12 , filament_ratio = (2, 3, 1),
                                 n_batchs = 5, transforms = transforms1, image_dim = 224)
```
```python
for _, imgs, labels in data_loader:
    print("size of images ",imgs.shape)
    print("labels for each batch ",labels)
```
    size of images  torch.Size([12, 224, 224])
    labels for each batch  tensor([[-1],
            [-1],
            [ 1],
            [-1],
            [ 0],
            [ 1],
            [-1],
            [ 1],
            [ 1],
            [ 1],
            [ 0],
            [ 1]])
    size of images  torch.Size([12, 224, 224])
    labels for each batch  tensor([[ 0],
            [ 1],
            [-1],
            [-1],
            [ 1],
            [-1],
            [ 1],
            [ 1],
            [ 0],
            [-1],
            [ 1],
            [ 1]])
    size of images  torch.Size([12, 224, 224])
    labels for each batch  tensor([[ 1],
            [ 1],
            [ 1],
            [ 0],
            [-1],
            [ 1],
            [-1],
            [ 0],
            [-1],
            [ 1],
            [-1],
            [ 1]])
    size of images  torch.Size([12, 224, 224])
    labels for each batch  tensor([[-1],
            [-1],
            [ 1],
            [ 1],
            [ 1],
            [ 0],
            [ 1],
            [-1],
            [-1],
            [ 1],
            [ 1],
            [ 0]])
    size of images  torch.Size([12, 224, 224])
    labels for each batch  tensor([[ 1],
            [ 1],
            [-1],
            [ 1],
            [-1],
            [ 0],
            [ 1],
            [ 0],
            [ 1],
            [-1],
            [-1],
            [ 1]])
#### How to generate 10 filament images for every batch only for left and right chirality for 5 batches with image dimension of 224x224 ?
- In order to remove one type of chiraity, filament ratio, i.e., tuple(L, R, U):   
    - if L=0 that means left chirality is eliminated. Similarly, this applies to other types as well.
```python
data_loader = FilamentDataLoader(dataset = dataset,batch_size = 10 , filament_ratio = (1, 1, 0),
                                 n_batchs = 5, transforms = transforms1, image_dim = 224)
```
```python
for _, imgs, labels in data_loader:
    print("size of images ",imgs.shape)
    print("labels for each batch ",labels)
```
    size of images  torch.Size([10, 224, 224])
    labels for each batch  tensor([[-1],
            [-1],
            [ 1],
            [ 1],
            [ 1],
            [-1],
            [ 1],
            [-1],
            [-1],
            [ 1]])
    size of images  torch.Size([10, 224, 224])
    labels for each batch  tensor([[ 1],
            [-1],
            [-1],
            [ 1],
            [-1],
            [-1],
            [ 1],
            [ 1],
            [ 1],
            [-1]])
    size of images  torch.Size([10, 224, 224])
    labels for each batch  tensor([[ 1],
            [ 1],
            [ 1],
            [ 1],
            [-1],
            [ 1],
            [-1],
            [-1],
            [-1],
            [-1]])
    size of images  torch.Size([10, 224, 224])
    labels for each batch  tensor([[ 1],
            [ 1],
            [-1],
            [-1],
            [ 1],
            [-1],
            [-1],
            [ 1],
            [-1],
            [ 1]])
    size of images  torch.Size([10, 224, 224])
    labels for each batch  tensor([[-1],
            [-1],
            [-1],
            [ 1],
            [ 1],
            [ 1],
            [-1],
            [ 1],
            [-1],
            [ 1]])

%package -n python3-augmentation-engine
Summary:	Solar Filaments data augmentation demo package
Provides:	python-augmentation-engine
BuildRequires:	python3-devel
BuildRequires:	python3-setuptools
BuildRequires:	python3-pip
%description -n python3-augmentation-engine
```python
pip install augmentation_engine
```
    Requirement already satisfied: augmentation_engine in d:\gsu_assignments\semester_2\dl\augmentation_engine (0.0.1)
    Requirement already satisfied: sortedcontainers in c:\users\shreejaa talla\pycharmprojects\bbso_fa\venv\lib\site-packages (from augmentation_engine) (2.4.0)
    Requirement already satisfied: opencv_python in c:\users\shreejaa talla\pycharmprojects\bbso_fa\venv\lib\site-packages (from augmentation_engine) (4.5.3.56)
    Requirement already satisfied: torchvision in c:\users\shreejaa talla\pycharmprojects\bbso_fa\venv\lib\site-packages (from augmentation_engine) (0.10.0)
    Requirement already satisfied: pillow in c:\users\shreejaa talla\pycharmprojects\bbso_fa\venv\lib\site-packages (from augmentation_engine) (8.3.2)
    Requirement already satisfied: numpy>=1.17.3 in c:\users\shreejaa talla\pycharmprojects\bbso_fa\venv\lib\site-packages (from opencv_python->augmentation_engine) (1.21.2)
    Requirement already satisfied: torch==1.9.0 in c:\users\shreejaa talla\pycharmprojects\bbso_fa\venv\lib\site-packages (from torchvision->augmentation_engine) (1.9.0)
    Requirement already satisfied: typing-extensions in c:\users\shreejaa talla\pycharmprojects\bbso_fa\venv\lib\site-packages (from torch==1.9.0->torchvision->augmentation_engine) (3.10.0.2)
    Note: you may need to restart the kernel to use updated packages.
### Import Required Libraries 
```python
import os
from torchvision import transforms
import matplotlib.pyplot as plt
from filament_augmentation.loader.filament_dataloader import FilamentDataLoader
from filament_augmentation.generator.filament_dataset import FilamentDataset
from filament_augmentation.metadata.filament_metadata import FilamentMetadata
```
**To find out the number of left, right and unidentified chiralities for an interval of time.**
- The code snippet below gives the *chirality distribution*, i.e., the distribution of left, right and unidentified chiralities for an interval of time from "2015-08-01 17:36:15" to "2015-08-09 18:15:17".
- Here the petdata has big bear space observatory(BBSO) full disk solar images from (01-09) aug 2015.
- The format for start and end time should be YYYY-MM-DD HH:MM:SS.
- The ann_file or annotation file is a H-alpha based manually labelled filaments in a json file comes with petdata.
```python
__file__ = 'augmentation_process.ipynb'
bbso_json = os.path.abspath(
        os.path.join(os.path.dirname(__file__), 'petdata', 'bbso_json_data','2015_chir_data.json'))
filamentInfo = FilamentMetadata(ann_file = bbso_json, start_time = '2015-08-01 00:00:15',
                                    end_time = '2015-08-30 23:59:59')
filamentInfo.get_chirality_distribution()
```
    (22, 30, 186)
- In order to generate extra filaments for left, right or unidentified chirality by either balancing the data or getting them in required ratios to train them using an ML algorithm. A filament dataset class should be initialized which is quite similar to that of pytorch dataset class.
- The transform list should be list of torchvision [transformations](https://pytorch.org/vision/0.8/transforms.html) 
- Filament ratio is tuple variable that takes (L,R,U).
### Initializing Filament dataset 
To initialize filament dataset class follow parameters are required:
- bbso_path - BBSO full disk H-alpha solar images comes with petdata, path of the folder.
- ann_file - a H-alpha based manually labelled filaments in a json file comes with petdata.
- The format for start and end time should be YYYY-MM-DD HH:MM:SS.
```python
bbso_path = os.path.abspath(os.path.join(os.path.dirname(__file__), 'petdata', '2015'))
dataset = FilamentDataset(bbso_path = bbso_path, ann_file = bbso_json, 
                          start_time = "2015-08-01 17:36:15", end_time = "2015-08-09 17:36:15")
```
    loading annotations into memory...
    Done (t=0.07s)
    creating index...
    index created!
### Setup transformations for data augmentation
The transformations function can be refered from [torchvision transforms](https://pytorch.org/vision/0.8/transforms.html)
- Here transforms variable should have list of torchvision transforms functions as shown below: 
```python
transforms1 = [
    transforms.ColorJitter(brightness=(0.25,1.25), contrast=(0.25,2.00), saturation=(0.25,2.25)),
    transforms.RandomRotation(15,expand=False,fill=110)
]
```
### Initializing data loader
- dataset = object of filament dataset class.
- batch_size = number of filaments to be generated per batch.
- filament_ratio = tuple of three values, i.e., ratios of left, right and unidentified chirality to be generated in each batch.
- n_batchs = number of batchs.
- transforms = list of torchvision transformations functions
- image_dim = image dimensions if image dimension is -1 then image will not be resize, i.e., output is original image size.
```python
data_loader = FilamentDataLoader(dataset = dataset,batch_size = 3 , filament_ratio = (1, 1, 1),n_batchs = 10, 
                                 transforms = transforms1, image_dim = 224)
```
#### How to generate 3 filament images for every batch with ratio of left as 1, right chirality as 1 and unidentified as 1 for 10 batches with original image dimension and display the images?
```python
data_loader = FilamentDataLoader(dataset = dataset,batch_size = 3 , filament_ratio = (1, 1, 1),
                                 n_batchs = 10, transforms = transforms1, image_dim = -1)
```
#### Batch -1 augmented filament images and their following labels (1=R, 0= U, -1=L)
```python
for original_imgs, transformed_imgs, labels in data_loader:
    for org_img, img, label in zip(original_imgs ,transformed_imgs, labels):
        print("Original image")
        plt.imshow(org_img, cmap='gray')
        plt.show()
        print("Transformed image")
        plt.imshow(img, cmap='gray')
        plt.show()
        print("Label",label)
    break
```
    Original image
![png](document_images/output_18_1.png)
    Transformed image
![png](document_images/output_18_3.png)
    Label 0
    Original image
![png](document_images/output_18_5.png)
    Transformed image
![png](document_images/output_18_7.png)
    Label 1
    Original image
![png](document_images/output_18_9.png)
    Transformed image
![png](document_images/output_18_11.png)
    Label -1
#### How to generate 12  filament images for every batch with ratio of left as 2, right chirality as 3  and unidentified as 1 for 5 batches with image dimension of 224x224 ?
```python
data_loader = FilamentDataLoader(dataset = dataset,batch_size = 12 , filament_ratio = (2, 3, 1),
                                 n_batchs = 5, transforms = transforms1, image_dim = 224)
```
```python
for _, imgs, labels in data_loader:
    print("size of images ",imgs.shape)
    print("labels for each batch ",labels)
```
    size of images  torch.Size([12, 224, 224])
    labels for each batch  tensor([[-1],
            [-1],
            [ 1],
            [-1],
            [ 0],
            [ 1],
            [-1],
            [ 1],
            [ 1],
            [ 1],
            [ 0],
            [ 1]])
    size of images  torch.Size([12, 224, 224])
    labels for each batch  tensor([[ 0],
            [ 1],
            [-1],
            [-1],
            [ 1],
            [-1],
            [ 1],
            [ 1],
            [ 0],
            [-1],
            [ 1],
            [ 1]])
    size of images  torch.Size([12, 224, 224])
    labels for each batch  tensor([[ 1],
            [ 1],
            [ 1],
            [ 0],
            [-1],
            [ 1],
            [-1],
            [ 0],
            [-1],
            [ 1],
            [-1],
            [ 1]])
    size of images  torch.Size([12, 224, 224])
    labels for each batch  tensor([[-1],
            [-1],
            [ 1],
            [ 1],
            [ 1],
            [ 0],
            [ 1],
            [-1],
            [-1],
            [ 1],
            [ 1],
            [ 0]])
    size of images  torch.Size([12, 224, 224])
    labels for each batch  tensor([[ 1],
            [ 1],
            [-1],
            [ 1],
            [-1],
            [ 0],
            [ 1],
            [ 0],
            [ 1],
            [-1],
            [-1],
            [ 1]])
#### How to generate 10 filament images for every batch only for left and right chirality for 5 batches with image dimension of 224x224 ?
- In order to remove one type of chiraity, filament ratio, i.e., tuple(L, R, U):   
    - if L=0 that means left chirality is eliminated. Similarly, this applies to other types as well.
```python
data_loader = FilamentDataLoader(dataset = dataset,batch_size = 10 , filament_ratio = (1, 1, 0),
                                 n_batchs = 5, transforms = transforms1, image_dim = 224)
```
```python
for _, imgs, labels in data_loader:
    print("size of images ",imgs.shape)
    print("labels for each batch ",labels)
```
    size of images  torch.Size([10, 224, 224])
    labels for each batch  tensor([[-1],
            [-1],
            [ 1],
            [ 1],
            [ 1],
            [-1],
            [ 1],
            [-1],
            [-1],
            [ 1]])
    size of images  torch.Size([10, 224, 224])
    labels for each batch  tensor([[ 1],
            [-1],
            [-1],
            [ 1],
            [-1],
            [-1],
            [ 1],
            [ 1],
            [ 1],
            [-1]])
    size of images  torch.Size([10, 224, 224])
    labels for each batch  tensor([[ 1],
            [ 1],
            [ 1],
            [ 1],
            [-1],
            [ 1],
            [-1],
            [-1],
            [-1],
            [-1]])
    size of images  torch.Size([10, 224, 224])
    labels for each batch  tensor([[ 1],
            [ 1],
            [-1],
            [-1],
            [ 1],
            [-1],
            [-1],
            [ 1],
            [-1],
            [ 1]])
    size of images  torch.Size([10, 224, 224])
    labels for each batch  tensor([[-1],
            [-1],
            [-1],
            [ 1],
            [ 1],
            [ 1],
            [-1],
            [ 1],
            [-1],
            [ 1]])

%package help
Summary:	Development documents and examples for augmentation-engine
Provides:	python3-augmentation-engine-doc
%description help
```python
pip install augmentation_engine
```
    Requirement already satisfied: augmentation_engine in d:\gsu_assignments\semester_2\dl\augmentation_engine (0.0.1)
    Requirement already satisfied: sortedcontainers in c:\users\shreejaa talla\pycharmprojects\bbso_fa\venv\lib\site-packages (from augmentation_engine) (2.4.0)
    Requirement already satisfied: opencv_python in c:\users\shreejaa talla\pycharmprojects\bbso_fa\venv\lib\site-packages (from augmentation_engine) (4.5.3.56)
    Requirement already satisfied: torchvision in c:\users\shreejaa talla\pycharmprojects\bbso_fa\venv\lib\site-packages (from augmentation_engine) (0.10.0)
    Requirement already satisfied: pillow in c:\users\shreejaa talla\pycharmprojects\bbso_fa\venv\lib\site-packages (from augmentation_engine) (8.3.2)
    Requirement already satisfied: numpy>=1.17.3 in c:\users\shreejaa talla\pycharmprojects\bbso_fa\venv\lib\site-packages (from opencv_python->augmentation_engine) (1.21.2)
    Requirement already satisfied: torch==1.9.0 in c:\users\shreejaa talla\pycharmprojects\bbso_fa\venv\lib\site-packages (from torchvision->augmentation_engine) (1.9.0)
    Requirement already satisfied: typing-extensions in c:\users\shreejaa talla\pycharmprojects\bbso_fa\venv\lib\site-packages (from torch==1.9.0->torchvision->augmentation_engine) (3.10.0.2)
    Note: you may need to restart the kernel to use updated packages.
### Import Required Libraries 
```python
import os
from torchvision import transforms
import matplotlib.pyplot as plt
from filament_augmentation.loader.filament_dataloader import FilamentDataLoader
from filament_augmentation.generator.filament_dataset import FilamentDataset
from filament_augmentation.metadata.filament_metadata import FilamentMetadata
```
**To find out the number of left, right and unidentified chiralities for an interval of time.**
- The code snippet below gives the *chirality distribution*, i.e., the distribution of left, right and unidentified chiralities for an interval of time from "2015-08-01 17:36:15" to "2015-08-09 18:15:17".
- Here the petdata has big bear space observatory(BBSO) full disk solar images from (01-09) aug 2015.
- The format for start and end time should be YYYY-MM-DD HH:MM:SS.
- The ann_file or annotation file is a H-alpha based manually labelled filaments in a json file comes with petdata.
```python
__file__ = 'augmentation_process.ipynb'
bbso_json = os.path.abspath(
        os.path.join(os.path.dirname(__file__), 'petdata', 'bbso_json_data','2015_chir_data.json'))
filamentInfo = FilamentMetadata(ann_file = bbso_json, start_time = '2015-08-01 00:00:15',
                                    end_time = '2015-08-30 23:59:59')
filamentInfo.get_chirality_distribution()
```
    (22, 30, 186)
- In order to generate extra filaments for left, right or unidentified chirality by either balancing the data or getting them in required ratios to train them using an ML algorithm. A filament dataset class should be initialized which is quite similar to that of pytorch dataset class.
- The transform list should be list of torchvision [transformations](https://pytorch.org/vision/0.8/transforms.html) 
- Filament ratio is tuple variable that takes (L,R,U).
### Initializing Filament dataset 
To initialize filament dataset class follow parameters are required:
- bbso_path - BBSO full disk H-alpha solar images comes with petdata, path of the folder.
- ann_file - a H-alpha based manually labelled filaments in a json file comes with petdata.
- The format for start and end time should be YYYY-MM-DD HH:MM:SS.
```python
bbso_path = os.path.abspath(os.path.join(os.path.dirname(__file__), 'petdata', '2015'))
dataset = FilamentDataset(bbso_path = bbso_path, ann_file = bbso_json, 
                          start_time = "2015-08-01 17:36:15", end_time = "2015-08-09 17:36:15")
```
    loading annotations into memory...
    Done (t=0.07s)
    creating index...
    index created!
### Setup transformations for data augmentation
The transformations function can be refered from [torchvision transforms](https://pytorch.org/vision/0.8/transforms.html)
- Here transforms variable should have list of torchvision transforms functions as shown below: 
```python
transforms1 = [
    transforms.ColorJitter(brightness=(0.25,1.25), contrast=(0.25,2.00), saturation=(0.25,2.25)),
    transforms.RandomRotation(15,expand=False,fill=110)
]
```
### Initializing data loader
- dataset = object of filament dataset class.
- batch_size = number of filaments to be generated per batch.
- filament_ratio = tuple of three values, i.e., ratios of left, right and unidentified chirality to be generated in each batch.
- n_batchs = number of batchs.
- transforms = list of torchvision transformations functions
- image_dim = image dimensions if image dimension is -1 then image will not be resize, i.e., output is original image size.
```python
data_loader = FilamentDataLoader(dataset = dataset,batch_size = 3 , filament_ratio = (1, 1, 1),n_batchs = 10, 
                                 transforms = transforms1, image_dim = 224)
```
#### How to generate 3 filament images for every batch with ratio of left as 1, right chirality as 1 and unidentified as 1 for 10 batches with original image dimension and display the images?
```python
data_loader = FilamentDataLoader(dataset = dataset,batch_size = 3 , filament_ratio = (1, 1, 1),
                                 n_batchs = 10, transforms = transforms1, image_dim = -1)
```
#### Batch -1 augmented filament images and their following labels (1=R, 0= U, -1=L)
```python
for original_imgs, transformed_imgs, labels in data_loader:
    for org_img, img, label in zip(original_imgs ,transformed_imgs, labels):
        print("Original image")
        plt.imshow(org_img, cmap='gray')
        plt.show()
        print("Transformed image")
        plt.imshow(img, cmap='gray')
        plt.show()
        print("Label",label)
    break
```
    Original image
![png](document_images/output_18_1.png)
    Transformed image
![png](document_images/output_18_3.png)
    Label 0
    Original image
![png](document_images/output_18_5.png)
    Transformed image
![png](document_images/output_18_7.png)
    Label 1
    Original image
![png](document_images/output_18_9.png)
    Transformed image
![png](document_images/output_18_11.png)
    Label -1
#### How to generate 12  filament images for every batch with ratio of left as 2, right chirality as 3  and unidentified as 1 for 5 batches with image dimension of 224x224 ?
```python
data_loader = FilamentDataLoader(dataset = dataset,batch_size = 12 , filament_ratio = (2, 3, 1),
                                 n_batchs = 5, transforms = transforms1, image_dim = 224)
```
```python
for _, imgs, labels in data_loader:
    print("size of images ",imgs.shape)
    print("labels for each batch ",labels)
```
    size of images  torch.Size([12, 224, 224])
    labels for each batch  tensor([[-1],
            [-1],
            [ 1],
            [-1],
            [ 0],
            [ 1],
            [-1],
            [ 1],
            [ 1],
            [ 1],
            [ 0],
            [ 1]])
    size of images  torch.Size([12, 224, 224])
    labels for each batch  tensor([[ 0],
            [ 1],
            [-1],
            [-1],
            [ 1],
            [-1],
            [ 1],
            [ 1],
            [ 0],
            [-1],
            [ 1],
            [ 1]])
    size of images  torch.Size([12, 224, 224])
    labels for each batch  tensor([[ 1],
            [ 1],
            [ 1],
            [ 0],
            [-1],
            [ 1],
            [-1],
            [ 0],
            [-1],
            [ 1],
            [-1],
            [ 1]])
    size of images  torch.Size([12, 224, 224])
    labels for each batch  tensor([[-1],
            [-1],
            [ 1],
            [ 1],
            [ 1],
            [ 0],
            [ 1],
            [-1],
            [-1],
            [ 1],
            [ 1],
            [ 0]])
    size of images  torch.Size([12, 224, 224])
    labels for each batch  tensor([[ 1],
            [ 1],
            [-1],
            [ 1],
            [-1],
            [ 0],
            [ 1],
            [ 0],
            [ 1],
            [-1],
            [-1],
            [ 1]])
#### How to generate 10 filament images for every batch only for left and right chirality for 5 batches with image dimension of 224x224 ?
- In order to remove one type of chiraity, filament ratio, i.e., tuple(L, R, U):   
    - if L=0 that means left chirality is eliminated. Similarly, this applies to other types as well.
```python
data_loader = FilamentDataLoader(dataset = dataset,batch_size = 10 , filament_ratio = (1, 1, 0),
                                 n_batchs = 5, transforms = transforms1, image_dim = 224)
```
```python
for _, imgs, labels in data_loader:
    print("size of images ",imgs.shape)
    print("labels for each batch ",labels)
```
    size of images  torch.Size([10, 224, 224])
    labels for each batch  tensor([[-1],
            [-1],
            [ 1],
            [ 1],
            [ 1],
            [-1],
            [ 1],
            [-1],
            [-1],
            [ 1]])
    size of images  torch.Size([10, 224, 224])
    labels for each batch  tensor([[ 1],
            [-1],
            [-1],
            [ 1],
            [-1],
            [-1],
            [ 1],
            [ 1],
            [ 1],
            [-1]])
    size of images  torch.Size([10, 224, 224])
    labels for each batch  tensor([[ 1],
            [ 1],
            [ 1],
            [ 1],
            [-1],
            [ 1],
            [-1],
            [-1],
            [-1],
            [-1]])
    size of images  torch.Size([10, 224, 224])
    labels for each batch  tensor([[ 1],
            [ 1],
            [-1],
            [-1],
            [ 1],
            [-1],
            [-1],
            [ 1],
            [-1],
            [ 1]])
    size of images  torch.Size([10, 224, 224])
    labels for each batch  tensor([[-1],
            [-1],
            [-1],
            [ 1],
            [ 1],
            [ 1],
            [-1],
            [ 1],
            [-1],
            [ 1]])

%prep
%autosetup -n augmentation-engine-0.1.2

%build
%py3_build

%install
%py3_install
install -d -m755 %{buildroot}/%{_pkgdocdir}
if [ -d doc ]; then cp -arf doc %{buildroot}/%{_pkgdocdir}; fi
if [ -d docs ]; then cp -arf docs %{buildroot}/%{_pkgdocdir}; fi
if [ -d example ]; then cp -arf example %{buildroot}/%{_pkgdocdir}; fi
if [ -d examples ]; then cp -arf examples %{buildroot}/%{_pkgdocdir}; fi
pushd %{buildroot}
if [ -d usr/lib ]; then
	find usr/lib -type f -printf "/%h/%f\n" >> filelist.lst
fi
if [ -d usr/lib64 ]; then
	find usr/lib64 -type f -printf "/%h/%f\n" >> filelist.lst
fi
if [ -d usr/bin ]; then
	find usr/bin -type f -printf "/%h/%f\n" >> filelist.lst
fi
if [ -d usr/sbin ]; then
	find usr/sbin -type f -printf "/%h/%f\n" >> filelist.lst
fi
touch doclist.lst
if [ -d usr/share/man ]; then
	find usr/share/man -type f -printf "/%h/%f.gz\n" >> doclist.lst
fi
popd
mv %{buildroot}/filelist.lst .
mv %{buildroot}/doclist.lst .

%files -n python3-augmentation-engine -f filelist.lst
%dir %{python3_sitelib}/*

%files help -f doclist.lst
%{_docdir}/*

%changelog
* Wed May 31 2023 Python_Bot <Python_Bot@openeuler.org> - 0.1.2-1
- Package Spec generated