Ecosystem/misc-research
1
0
Fork 0
mirror of https://github.com/Rucknium/misc-research.git synced 2024-11-16 15:58:03 +00:00
Code Issues Projects Releases Packages Wiki Activity Actions

Revert prior optimization commit due to troubleshooting failure

Browse source
This commit is contained in:
Rucknium 2022-08-09 22:53:08 +00:00
parent f613b58ea7
commit 912f221f68
1 changed files with 9 additions and 20 deletions
Show stats Download patch file Download diff file Expand all files Collapse all files

29
General-Blockchain-Age-of-Spent-Outputs/construct-edgelist.R
View file

@ -9,6 +9,10 @@ library(DBI)
data.dir <- "" data.dir <- ""
# Input data directory here, with trailing "/" # Input data directory here, with trailing "/"
source("https://gist.githubusercontent.com/jeffwong/5925000/raw/bf02ed0dd2963169a91664be02fb18e45c4d1e20/sqlitewritetable.R")
# From https://gist.github.com/jeffwong/5925000
# Modifies RSQLite's sqliteWriteTable function so as to reject duplicates
con <- DBI::dbConnect(RSQLite::SQLite(), paste0(data.dir, "tx-graph-node-indices.db")) con <- DBI::dbConnect(RSQLite::SQLite(), paste0(data.dir, "tx-graph-node-indices.db"))
DBI::dbExecute(con, "PRAGMA max_page_count = 4294967292;") DBI::dbExecute(con, "PRAGMA max_page_count = 4294967292;")
@ -21,7 +25,8 @@ DBI::dbExecute(con, "PRAGMA max_page_count = 4294967292;")
DBI::dbExecute(con, "CREATE TABLE nodes ( DBI::dbExecute(con, "CREATE TABLE nodes (
node TEXT, node TEXT,
node_index INTEGER PRIMARY KEY node_index INTEGER PRIMARY KEY AUTOINCREMENT,
unique(node)
)") )")
@ -63,31 +68,15 @@ for (file.iter in tx.graph.files) {
new.nodes <- unique(c(tx.graph.chunk$origin, tx.graph.chunk$destination)) new.nodes <- unique(c(tx.graph.chunk$origin, tx.graph.chunk$destination))
nodes.to.insert <- data.frame(node = new.nodes, stringsAsFactors = FALSE) nodes.to.insert <- data.frame(node = new.nodes, node_index = NA, stringsAsFactors = FALSE)
DBI::dbWriteTable(con, "nodes",
nodes.to.insert, append = TRUE)
mysqliteWriteTable(con, "nodes",
nodes.to.insert, append = TRUE, row.names = FALSE, ignore = TRUE)
cat(nrow(nodes.to.insert), "Nodes written\n") cat(nrow(nodes.to.insert), "Nodes written\n")
} }
base::date()
DBI::dbExecute(con, "DELETE FROM nodes
WHERE rowid NOT IN (
SELECT MIN(rowid)
FROM nodes
GROUP BY node
)")
base::date()
# 1 hour
# Suggested by https://stackoverflow.com/questions/25884095/how-can-i-delete-duplicates-in-sqlite
# Must do this at the end to make sure did not leave any out
# from the tail end of the last 1000 iters
DBI::dbWriteTable(con, "edgelist_intermediate_1", DBI::dbWriteTable(con, "edgelist_intermediate_1",
data.frame(origin = character(0), destination = character(0), data.frame(origin = character(0), destination = character(0),