subspaceGeneration: subspaceGeneration

In alejandrommingo/QLSA: Quantum Latent Semantic Analysis

subspaceGeneration

Description

subspaceGeneration is a function that decides the number of dimensions a contour deserves using the 'paran' package, compute an EFA (Exploratory Factor Analysis) with that number of dimensions using the 'psych' package, and reorthogonalize the solution to define a basis for the new subspace using the 'pracma' package.

Usage

subspaceGeneration(
  word,
  gallitoCode,
  spaceName,
  min_cosine = 0.5,
  min_reilability = 0.85,
  min_cosine_contour = 0.3,
  neighbors = 100
)

Arguments

word	A string or a tupple of strings separated by "_" indicating the word for which you want to define the subspace.
gallitoCode	Gallito API password to extract information from the LSA semantic space.
spaceName	Gallito API LSA semantic space you want to use.
min_cosine	The minimum cosine the function will use to return the plots of similar words for each new dimension of the subspace. By default 'min_cosine = 0.5'.
min_reilability	The minimum reilability the function will consider to decide that a reorthogonalized dimension is equivalent to the dimension extracted in the factorial solution. By default 'min_reilability = 0.85'.
min_cosine_contour	minimum cosine required for being a contour exemplar. By default 'min_cosine_contour = 0.3'.
neighbors	The number of neighbors inside the contour of the word. By default 'neighbors = 100'.

Value

The function will return a list with the subspace as 'subspace', the reilability test as 'reilability_test', the subspace graphical information as 'subspace_info' and the EFA results as 'EFA_info'.

alejandrommingo/QLSA documentation built on March 19, 2022, 3:36 a.m.