Using grep to filter rows with two or more patterns in the string in R

Question

I need to index all the rows that have a string beginning with either "B-" or "B^" in one of the columns. I tried a bunch of combinations, but I am suspecting it might not be working due to "-" and "^" signs being part of grep command as well.

dataset[grep('^(B-|B^)[^B-|B^]*$', dataset$Col1),]

With the above script, rows beginning with "B^" are not being extracted. Please suggest a smart way to handle this.

Answer 1

您可以在grep使用转义\\\\命令：

dataset[grep('^(B\\-|B\\^)[^B\\-|B\\^]*$', dataset$Col1),]

Answer 2

For further explanation, the ^ matches the beginning of a string as an anchor therefore you have to escape it in the middle of string. The [] are a character class so [^B-|B^]* matches any character that's not a B,-,B, or ^. They are unnecessary here.

The simplified regex is: dataset[grep('^(B-|B\\\\^)', dataset$Col1),]