PySpark EDA & Datasets Conversion

- Must make use of PySpark for the Exploratory Data Analysis. Do not have to train model using PySpark
- Add comments / describe the EDA
- Need to convert from PySpark to Pandas for the test train split.
- And load into PySpark

- Dataset Hyperparameters:
* Forced images to be 48x48
* Using PyTorch not IntensiveLock

I can send you the notebook so far.
You just need to get it to the point where PySpark data frame converts to pandas and then the train test split before training.
If you use online resources then just copy and paste the links in the relevant notebook cells

W.r.t. the convolutional neural network it is basic.
Pooling layer
The convolutional layer
Nothing hardset, flexible at the moment and nothing computationally complex

Pyton - 3.9.x
Java (JDK) - 8
PySpark - 3.3.2
NumPy - 1.26.4
pandas - 1.5.3
PyArrow - 10.0.1
Variables - set manually in Command Prompt before Spark

Date Requirement: Please make the labels into even distribution of 5000 each ( ONLY HAPPY and ANGRY) at the least ( min 10000 rows of data)

Dataset: https://www.kaggle.com/datasets/samithsachidanandan/human-face-emotions