read.table和readr::read

# 1. 使用read.table读取数据 system.time( read.table("/home/data/test_data", sep = "\001", quote = "", stringsAsFactors = F, comment.char = "", col.names = colNames) ) # colNames为预先定义的列名； # 也可以设置为 ：col.names = TRUE / FALSE # user system elapsed # 67.943 0.277 68.326 # 2.使用readr::read_delim读取数据 library(readr) system.time( read_delim("/home/data/test_data", delim = "\001", quote = "", comment = "", col_names = colNames) ) # colNames为预先定义的列名； # 也可以设置为 ：col.names = TRUE / FALSE # =================================| 100% 796 MB # user system elapsed # 12.790 0.245 12.947

可以看出，读取796MB的数据test_data,read.table所用时间为67.943s,而read_delim只需要12.790s;读取速度有显著的提升，大约为read.table的5倍。