Pandas DataFrame as input

[adamov-artem]

Hi. Is it possible to run geosketch on a DataFrame object? I need to save row- and colnames of my count matrix.

[brianhie]

@adamov-artem you can call to_numpy() to get the count matrix, then do the PCA/geosketch routine, and then use the indices returned by gs() to index into your data frame using iloc().

For example,

X = data_frame.to_numpy()

# PCA.
from fbpca import pca
U, s, Vt = pca(X, k=100) # E.g., 100 PCs.
X_dimred = U[:, :100] * s[:100]

# Sketch.
from geosketch import gs
N = 20000 # Number of samples to obtain from the data set.
sketch_index = gs(X_dimred, N, replace=False)

data_frame_sketch = data_frame.iloc[sketch_index, :]