Expand network through iterative growth

Expands an initial network by iteratively finding new users who are followed by many existing network members and match specified keywords. This is the core algorithm for growing networks from a seed set of users.

Usage

expand_net(
  net,
  keywords,
  token,
  refresh_tok,
  save_net = FALSE,
  file_name,
  threshold,
  max_iterations,
  sample_size = Inf
)

Arguments

net

Tibble. Initial network with follow relationships, must contain:

actor_handle

Character. User who is following

follows_handle

Character. User being followed

keywords

Character vector. Keywords to search for in user profile descriptions to determine network membership eligibility

token

Character. Authentication token from get_token

refresh_tok

Character. Refresh token from get_token for long-running operations

save_net

Logical. Should the network be saved incrementally during expansion to prevent data loss? (default FALSE)

file_name

Character. File path for incremental saves. If not provided, defaults to 'dev/net/bignet_TIMESTAMP.rds'

threshold

Numeric. Inclusion threshold for new users:

>= 1

Minimum number of existing network members who must follow a prospect

< 1

Fraction of network members who must follow a prospect

max_iterations

Integer. Maximum number of expansion iterations (default 50)

sample_size

Integer. Maximum prospects to consider per iteration. Use Inf for no limit (default), smaller values for testing

Value

A list containing the expanded network results:

net

Tibble. Expanded network with all follow relationships discovered

token

Character. Updated access token (may have been refreshed)

refresh_tok

Character. Updated refresh token

profiles

Tibble. Profile information for all network members

iterations

Integer. Number of iterations completed

See also

init_net, build_network, trim_net

Other network-building: build_network(), init_net(), trim_net()

Examples

if (FALSE) { # \dontrun{
# Start with an initial network
auth <- get_token("your.handle.bsky.social", "your-app-password")
token <- auth$accessJwt
refresh_tok <- auth$refreshJwt

# Create initial network
initial <- init_net("example.scientist.bsky.social",
                    c("research", "science"), token)

# Expand the network - conservative approach
expanded <- expand_net(
  net = initial,
  keywords = c("research", "science", "academic", "PhD"),
  token = token,
  refresh_tok = refresh_tok,
  threshold = 5,        # Need 5+ followers from existing network
  max_iterations = 10,  # Limit iterations for testing
  save_net = TRUE       # Save progress incrementally
)

print(paste("Network grew from", nrow(initial), "to",
            nrow(expanded$net), "connections"))

# More aggressive expansion with lower threshold
large_expansion <- expand_net(
  net = expanded$net,
  keywords = c("research", "science", "academic"),
  token = expanded$token,
  refresh_tok = expanded$refresh_tok,
  threshold = 0.1,      # 10% of network must follow prospect
  max_iterations = 25
)
} # }