{ "id": 1144, "title": "I gave up. Maybe this will help.", "url": "/competitions/WhatDoYouKnow/discussion/1144", "author_name": "", "post_date": "2011-12-10T17:52:55.843Z", "votes": 1, "comment_count": 1, "views": 2144, "content": "
\r\n

It takes me so long to seperate the user_strength so I gave up. As a stufdent, I don't have that much of time. I have already tried to manipulate my algorithm to do faster calculation. But it still is a very time consuming process. I post it and if anyone\r\n is interested, just copy it. By the, it is in Matlab. Userid is for valid_train and arranged according to the sequence that user id appear in the dataset. Anyone prefer csv can just change file name to csv. Codes are in para.txt and userid is in userid.txt.

\r\n

Hope you guys can do great.

\r\n

Bill Wang

\r\n

 

\r\n
", "messages": [ { "id": "7060", "postDate": "12/10/2011 17:52:55", "content": "
\r\n

It takes me so long to seperate the user_strength so I gave up. As a stufdent, I don't have that much of time. I have already tried to manipulate my algorithm to do faster calculation. But it still is a very time consuming process. I post it and if anyone\r\n is interested, just copy it. By the, it is in Matlab. Userid is for valid_train and arranged according to the sequence that user id appear in the dataset. Anyone prefer csv can just change file name to csv. Codes are in para.txt and userid is in userid.txt.

\r\n

Hope you guys can do great.

\r\n

Bill Wang

\r\n

 

\r\n
", "rawMarkdown": "", "votes": null }, { "id": "7062", "postDate": "12/10/2011 17:57:50", "content": "

I don't know why, but files are not attached.
\r\nSo code :

\r\n

data = load('el.txt');
\r\ny = data(:,1);
\r\nX = data(:,3:end);
\r\nuserid = load ('userid.txt');
\r\nnum = max(X(:,2));
\r\nnum = num+1;
\r\nuserid = [userid zeros(length(userid),2)];
\r\nquestionid = zeros(num,3);
\r\nfor k = 1 :num,
\r\nquestionid(k,1) = k-1;
\r\nend
\r\n[m n] = size(X);
\r\ni = 1;
\r\nfor k = 1:m,
\r\nt = userid(i,1) - X(1,1);
\r\nif t == 0,
\r\nuserid(i,2) = userid(i,2) + y(1,1);
\r\nuserid(i,3) = userid(i,3) + 1;
\r\nX = X(2:end,:);
\r\ny = y(2:end,:);
\r\nend
\r\nif t ~= 0,%% is not equal to
\r\ni = i+1;
\r\nuserid(i,2) = userid(i,2) + y(1,1);
\r\nuserid(i,3) = userid(i,3) + 1;
\r\nX = X(2:end,:); %% shorten the file and calculate faster
\r\ny = y(2:end,:);
\r\nend
\r\nk = k+1;
\r\nend

\r\n

for i = 1 : num,
\r\nfor q = 1 : length(X),
\r\nif question(i,1) - X(q,2) == 0,
\r\nquestionid(i,2) = question(i,2) + y(q,1);
\r\nquestionid(i,3) = question(i,3) + 1;
\r\nend
\r\nend
\r\nend

", "rawMarkdown": "", "votes": null } ], "comments": [ { "id": 7062, "author_name": "", "author_url": "", "post_date": "12/10/2011 17:57:50", "content": "

I don't know why, but files are not attached.
\r\nSo code :

\r\n

data = load('el.txt');
\r\ny = data(:,1);
\r\nX = data(:,3:end);
\r\nuserid = load ('userid.txt');
\r\nnum = max(X(:,2));
\r\nnum = num+1;
\r\nuserid = [userid zeros(length(userid),2)];
\r\nquestionid = zeros(num,3);
\r\nfor k = 1 :num,
\r\nquestionid(k,1) = k-1;
\r\nend
\r\n[m n] = size(X);
\r\ni = 1;
\r\nfor k = 1:m,
\r\nt = userid(i,1) - X(1,1);
\r\nif t == 0,
\r\nuserid(i,2) = userid(i,2) + y(1,1);
\r\nuserid(i,3) = userid(i,3) + 1;
\r\nX = X(2:end,:);
\r\ny = y(2:end,:);
\r\nend
\r\nif t ~= 0,%% is not equal to
\r\ni = i+1;
\r\nuserid(i,2) = userid(i,2) + y(1,1);
\r\nuserid(i,3) = userid(i,3) + 1;
\r\nX = X(2:end,:); %% shorten the file and calculate faster
\r\ny = y(2:end,:);
\r\nend
\r\nk = k+1;
\r\nend

\r\n

for i = 1 : num,
\r\nfor q = 1 : length(X),
\r\nif question(i,1) - X(q,2) == 0,
\r\nquestionid(i,2) = question(i,2) + y(q,1);
\r\nquestionid(i,3) = question(i,3) + 1;
\r\nend
\r\nend
\r\nend

", "votes": null, "replies": [] } ], "raw_markdown_by_id": { "7060": "", "7062": "" }, "source": "meta" }