Tanisha Patil, Per 3, Overall Individual Progress 2023-2024

image

Main works

Pocket Therapist

WW3 Mini Project

Personal Project

  • This is a model I created to forecast power voltage generation of a solar panel based on imaging data of weather conditions. It is a deep learning model I am building in pytorch, and an expansion on a very vague research paper done by professors at Stanford and Upenn.
  • I have background in AI and machine learning but when I was preprocessing my data, I decided to utilize a class based approach to hone in my skills from this class into my own project. So even though my model is HUGE (13 million parameters as of now) it is able to utilize getter functions that I wrote efficiently
  • I do not have it committed yet because of liscencing issues I am trying to navigate, but here is some code

class ImageDataset(Dataset): def __init__(self, h5_file, group, transform=None, target_transform=None): self.h5_file = h5_file self.group = group self.transform = transform self.target_transform = target_transform

    with h5py.File(h5_file, 'r') as f:
        images = f[self.group]['images_log'][:]
        labels = f[self.group]['pv_log'][:]

    self.images = images
    self.labels = labels

def __len__(self):
    return len(self.images)

def convert_to_float32(self, image, label):
    image = ((image - image.min()) / (image.max() - image.min()) * 255).astype(np.uint8)
    return image.astype(np.float32), label.astype(np.float32)

def __getitem__(self, idx):
    image = self.images[idx]
    label = self.labels[idx]

    # Convert to float32
    image, label = self.convert_to_float32(image, label)

    if self.transform:
        image = (image * 255).astype(np.uint8)
        image = self.transform(image)

    if self.target_transform:
        label = self.target_transform(label)

    return image, label </code>

Extra In Depth College Board

  • Old Blog LINK ~ 6 min read
  • Redid Blog to be more in Depth LINK ~ 24 min read
  • Target lesson on question 18 LINK

Overall

Commits

image

Reflection (Bulleted because who wants to read a paragraph)

  • I need to commit more sporadically
  • Work on collaboration and task assignment. Take on the role of scrum master and help my team stay on track
  • I want to learn automated deployment more in depth (why deploy with the same commands over and over again if I can just learn how to make a deployment script)
  • I want to work with more complicated APIs like AWS or Github to do some interesting data analysis
  • Maybe something like commits/contributions to predict success on live reviews
  • I want to try crowd sourcing data for atleast one project this year
  • I want to update my personal website with any cool projects I do in this class
  • I really want to do the student linkden project on the side
  • I would like to push myself to explore UI, since I spent most of this year focused on using Spring/Flask/Django for backend