{ "id": 3614, "title": "Reading in event_friends.csv into R", "url": "/competitions/event-recommendation-engine-challenge/discussion/3614", "author_name": "", "post_date": "2013-01-17T01:30:58.943Z", "votes": 1, "comment_count": 2, "views": 1376, "content": "

I use R

\r\n

I have trouble reading in a.o. the user_friends.csv file. If I use read.csv() I get two colums: One of type numerical and the other of type character. The first one is the user-id, the second is the list of all her friends. I use scan() to split up the second\r\n colum into individual user id's but that takes a very significant amount of time.

\r\n

Does anybody know a faster way? Can you post a snippet of code?

\r\n

Thanks

", "messages": [ { "id": "19379", "postDate": "01/17/2013 01:30:58", "content": "

I use R

\r\n

I have trouble reading in a.o. the user_friends.csv file. If I use read.csv() I get two colums: One of type numerical and the other of type character. The first one is the user-id, the second is the list of all her friends. I use scan() to split up the second\r\n colum into individual user id's but that takes a very significant amount of time.

\r\n

Does anybody know a faster way? Can you post a snippet of code?

\r\n

Thanks

", "rawMarkdown": "", "votes": null }, { "id": "19385", "postDate": "01/17/2013 07:26:57", "content": "

 

\r\n

\r\nHi\r\n Richard.

\r\n

\r\nMaybe\r\n you will be useful to this code snippet:

\r\n

\r\n#\r\n Read data

\r\n

\r\ndata<-read.csv(\r\n "user_friends.csv",\r\n header=TRUE,  sep=","\r\n ,stringsAsFactors\r\n = FALSE)

\r\n

\r\ndata$friends<-as.character(data$friends)

\r\n

\r\ndata$friends<-strsplit(data$friends,'\r\n ')

\r\n

\r\n#\r\n Determination of the number of friends, for the first entry

\r\n

\r\nnumber_friends<-length(data$friends[[1]])

\r\n

\r\n#\r\n Reading the index first friend

\r\n

\r\nind_friends_first<-as.numeric(data$friends[[1]][1])

\r\n

\r\n#\r\n Reading the index of the last friend

\r\n

\r\nind_friends_last<-as.numeric(data$friends[[1]][number_friends])

\r\n

\r\n>\r\n number_friends

\r\n

\r\n[1]\r\n 3623

\r\n

\r\n>\r\n ind_friends_first

\r\n

\r\n[1]\r\n 1346449342

\r\n

\r\n>\r\n ind_friends_last

\r\n

\r\n[1]\r\n 2409895224

\r\n

\r\n All the best,

\r\n

\r\nAlex

\r\n

\r\n 

", "rawMarkdown": "", "votes": null }, { "id": "19388", "postDate": "01/17/2013 08:25:29", "content": "

Alex,

\r\n

Thanks. I tried your suggestion: It works just fine. Still takes ~5 min before it is done, but much better than my solution (~1+ hr)

", "rawMarkdown": "", "votes": null } ], "comments": [ { "id": 19385, "author_name": "alexxanderlarko", "author_url": "", "post_date": "01/17/2013 07:26:57", "content": "

 

\r\n

\r\nHi\r\n Richard.

\r\n

\r\nMaybe\r\n you will be useful to this code snippet:

\r\n

\r\n#\r\n Read data

\r\n

\r\ndata<-read.csv(\r\n "user_friends.csv",\r\n header=TRUE,  sep=","\r\n ,stringsAsFactors\r\n = FALSE)

\r\n

\r\ndata$friends<-as.character(data$friends)

\r\n

\r\ndata$friends<-strsplit(data$friends,'\r\n ')

\r\n

\r\n#\r\n Determination of the number of friends, for the first entry

\r\n

\r\nnumber_friends<-length(data$friends[[1]])

\r\n

\r\n#\r\n Reading the index first friend

\r\n

\r\nind_friends_first<-as.numeric(data$friends[[1]][1])

\r\n

\r\n#\r\n Reading the index of the last friend

\r\n

\r\nind_friends_last<-as.numeric(data$friends[[1]][number_friends])

\r\n

\r\n>\r\n number_friends

\r\n

\r\n[1]\r\n 3623

\r\n

\r\n>\r\n ind_friends_first

\r\n

\r\n[1]\r\n 1346449342

\r\n

\r\n>\r\n ind_friends_last

\r\n

\r\n[1]\r\n 2409895224

\r\n

\r\n All the best,

\r\n

\r\nAlex

\r\n

\r\n 

", "votes": null, "replies": [] }, { "id": 19388, "author_name": "richardfromholland", "author_url": "", "post_date": "01/17/2013 08:25:29", "content": "

Alex,

\r\n

Thanks. I tried your suggestion: It works just fine. Still takes ~5 min before it is done, but much better than my solution (~1+ hr)

", "votes": null, "replies": [] } ], "raw_markdown_by_id": { "19379": "", "19385": "", "19388": "" }, "source": "meta" }