Feature Report: W&B Embeddings Projector

W&B's Embedding Projector allows users to plot multi-dimensional embeddings on a 2D plane using common dimension reduction algorithms like PCA, UMAP, and t-SNE.

Tim Sweeney

Created on December 1|Last edited on January 11

Comment

﻿
IntroEmbeddings are used to represent objects (people, images, posts, words, etc...) with a list of numbers - sometimes referred to as a vector. In machine learning and data science use cases, embeddings can be generated using a variety of approaches across a range of applications. This page assumes the reader is familiar with embeddings and is interested in visually analyzing them inside of W&B.
﻿Read the full docs here!
Code## https://scikit-learn.org/stable/datasets/toy_dataset.html
﻿
import wandb
from sklearn.datasets import load_iris, load_diabetes, load_digits, load_wine, load_breast_cancer
﻿
def get_df_from_sklearn_dataset(loader_fn):
  ds = loader_fn(as_frame=True)
  df = ds.data
  df["target"] = ds.target
  cols = df.columns.tolist()
  df = df[cols[-1:] + cols[:-1]]
  if (loader_fn == load_digits):
    df["image"] = df.apply(lambda row: wandb.Image(row[1:].values.reshape(8, 8) / 16.0), axis=1)
    cols = df.columns.tolist()
    df = df[cols[-1:] + cols[:-1]]
  return df
﻿
def get_all_dfs():
  return {name: get_df_from_sklearn_dataset(fn) for name, fn in ({
      "iris": load_iris,
      "diabetes": load_diabetes,
      "digits": load_digits,
      "wine": load_wine,
      "breast_cancer": load_breast_cancer,
  }).items()}
﻿
wandb.init(project="toy_datasets")
wandb.log(get_all_dfs())
wandb.finish()
Iris Dataset (150 records x 4 dimension - 3 Class Classification)
Wine Dataset (178 records x 13 dimensions - 3 Class Classification)
Diabetes Dataset (442 records x 10 dimensions - Regression)
Breast Cancer Dataset (569 records x 30 dimensions - Binary Classification)
Digits Dataset (1797 records x (32x32) dimensions - 10 Class Classification w/ Media)﻿

Add a comment