diff --git a/README.md b/README.md
index bd87b0a..d8522a3 100644
--- a/README.md
+++ b/README.md
@@ -134,3 +134,4 @@ The data has been sourced from repository at huggingface (https://huggingface.co
url="https://github.com/AI-Lab-Makerere/ibean/"
}
```
+
diff --git a/requirements.txt b/requirements.txt
index aa91466..80aefc0 100644
--- a/requirements.txt
+++ b/requirements.txt
@@ -19,4 +19,10 @@ pydantic~=1.10.13
flask~=3.0.0
python-dotenv~=1.0.0
pandas~=2.0.3
-mlflow~=2.8.0
\ No newline at end of file
+mlflow~=2.8.0
+deepchecks==0.17.5
+pytorch-ignite==0.4.13
+opencv-python==4.8.1.78
+scikit-image==0.22.0
+
+
diff --git a/src/data/deepchecks_validations.py b/src/data/deepchecks_validations.py
new file mode 100644
index 0000000..3e736cc
--- /dev/null
+++ b/src/data/deepchecks_validations.py
@@ -0,0 +1,63 @@
+
+import os
+import shutil
+from datasets import load_dataset
+from deepchecks.vision import classification_dataset_from_directory
+from deepchecks.vision.suites import train_test_validation
+
+# loads dataset for particular split
+def loadDataSet(name, split):
+ return load_dataset(name, split=split)
+
+# return the path of the cache images for particular split
+def pathToCopy(src, token):
+ pathList = src.split(os.sep)
+ result = []
+ while len(pathList) > 0:
+ if pathList[0] == token:
+ break
+ result.append(pathList.pop(0))
+
+ return os.path.join('/', *result)
+
+# copy images to data/raw
+def copyToRaw(src, out):
+ shutil.copytree(src, out, dirs_exist_ok=True)
+
+# prepare data for the analysis
+def prepareData(dsName, split, out):
+ # load test data
+ ds = loadDataSet(dsName, split)
+ # get the path of the cashed images
+ dsPath = pathToCopy(ds[0]["image_file_path"], split)
+ print(dsPath)
+ # copy test images from cache to raw folder
+ copyToRaw(dsPath, out)
+
+# conduct the analysis
+def runTrainTestValidation(src):
+ train, test = classification_dataset_from_directory(
+ root=src,
+ object_type='VisionData', image_extension='jpg')
+
+ suite = train_test_validation()
+ result = suite.run(train_dataset=train, test_dataset=test)
+ result.save_as_html('output.html')
+ print("TEST DONE!")
+
+# clear the data used for the analysis
+def clearData(src):
+ shutil.rmtree(src)
+
+def main():
+ datasetName ="beans"
+ outPath = os.path.join('data', 'raw')
+ prepareData(datasetName, 'train', outPath)
+ prepareData(datasetName, 'test', outPath)
+ runTrainTestValidation(outPath)
+ clearData(outPath)
+
+
+if __name__ == '__main__':
+ main()
+
diff --git a/src/data/output.html b/src/data/output.html
new file mode 100644
index 0000000..c3d0367
--- /dev/null
+++ b/src/data/output.html
@@ -0,0 +1,635 @@
+
+
+
+
+
+ Train Test Validation Suite
+
+
+
+
+
+
+
+