Calculating the number of Principal Components


Suppose there are 50 plant samples (rows) and 200 characteristics indices (columns). What will be the maximum number of Principal Components with positive eigenvalues we can get from such a dataset?

Based on my understanding of PCA, since 50 rows with 50 unit vectors and eigenvalues provide unique solutions, the maximum number of principal components should be 50.

Any corrections/suggestions/resources for this one?


© Copyright 2013-2020 Analytics Vidhya