[gov2001-l] Sampling rows - Gov2001 - lists.gking.harvard.edu

bfraga＠fas.harvard.edu

6 Feb 6 Feb

11:38 p.m.

I think it's just that you're missing the last step, which is to associate the draw of "row numbers" with the data for each of those rows. Code like the following might be helpful: yoursample <- samplespace[thedraw,] I haven't checked this, but hopefully this is a start. Good luck! -Bernard ----------------------- Bernard L. Fraga Ph.D. Student, Harvard University Government and Social Policy bfraga at fas.harvard.edu ----------------------- On Feb 6, 2009, at 11:13 PM, John-Paul Ferguson wrote:

...

Hi Everyone, I decided to get an early start on the problem set. I have a question about the sample() command, if anyone has gone that far (this relates to 2.4). In question 2.4, we are told to "Randomly draw 100 datasets with 50 observations each from the data loaded in part 2.2 with replacement...using sample()". I can't figure out how to get sample() to draw rows rather than single values. That is, if I write samplespace <- matrix(c(y,x1,x2,x3),1000,4) thedraw <- sample(samplespace, 50, replace=TRUE) then I get fifty scalars rather than fifty rows from the matrix samplespace. Can anyone see what I'm getting wrong here? The help on sample() isn't actually much help; it too seems to assume that you'll be drawing numbers. I normally wouldn't post such a "basic" question, but I figure that I can't be the only person who's puzzling out R syntax this week! Best, John-Paul Ferguson PhD Candidate, Economic Sociology MIT Sloan School of Management 50 Memorial Drive, E52-533 Cambridge, MA 02142 617.253.3940 (w) 617.549.8482 (c) http://web.mit.edu/jpferg/www _______________________________________________ gov2001-l mailing list gov2001-l at lists.fas.harvard.edu http://lists.fas.harvard.edu/mailman/listinfo/gov2001-l

Reply

jpferg＠mit.edu

12:58 a.m.

It's more the first step than the last--but yes, you're right. Thanks! --JP On Fri, Feb 6, 2009 at 11:38 PM, Bernard L. Fraga <bfraga at fas.harvard.edu>wrote:

...

I think it's just that you're missing the last step, which is to associate the draw of "row numbers" with the data for each of those rows. Code like the following might be helpful: yoursample <- samplespace[thedraw,] I haven't checked this, but hopefully this is a start. Good luck! -Bernard ----------------------- Bernard L. Fraga Ph.D. Student, Harvard University Government and Social Policy bfraga at fas.harvard.edu ----------------------- On Feb 6, 2009, at 11:13 PM, John-Paul Ferguson wrote: Hi Everyone, I decided to get an early start on the problem set. I have a question about the sample() command, if anyone has gone that far (this relates to 2.4). In question 2.4, we are told to "Randomly draw 100 datasets with 50 observations each from the data loaded in part 2.2 with replacement...using sample()". I can't figure out how to get sample() to draw *rows* rather than single *values*. That is, if I write samplespace <- matrix(c(y,x1,x2,x3),1000,4) thedraw <- sample(samplespace, 50, replace=TRUE) then I get fifty scalars rather than fifty rows from the matrix samplespace. Can anyone see what I'm getting wrong here? The help on sample() isn't actually much help; it too seems to assume that you'll be drawing numbers. I normally wouldn't post such a "basic" question, but I figure that I can't be the only person who's puzzling out R syntax this week! Best, John-Paul Ferguson PhD Candidate, Economic Sociology MIT Sloan School of Management 50 Memorial Drive, E52-533 Cambridge, MA 02142 617.253.3940 (w) 617.549.8482 (c) http://web.mit.edu/jpferg/www _______________________________________________ gov2001-l mailing list gov2001-l at lists.fas.harvard.edu http://lists.fas.harvard.edu/mailman/listinfo/gov2001-l _______________________________________________ gov2001-l mailing list gov2001-l at lists.fas.harvard.edu http://lists.fas.harvard.edu/mailman/listinfo/gov2001-l

Reply

jpferg＠mit.edu

9:08 a.m.

That is what I ultimately did last night, Diane, with Bernard's idea as the jumping-off point. I'd post the code but I'm not sure whether that's unsporting. JP On Sat, Feb 7, 2009 at 8:42 AM, Diane Kuhn <dmkuhn08 at gmail.com> wrote:

...

I think the way to think about this might be to draw a random sample of numbers from the numbers of observations in the data (so say to draw a random sample of numbers from the numbers 1-1000), and then once you have your random sample of numbers to call out/subset the dataset with the numbers from your sample: my.sample<-sample(x=1:nrow(data), size=samp.size, replace=TRUE) data.matrix<-data[my.sample,] hope that helps. -diane On Sat, Feb 7, 2009 at 12:58 AM, John-Paul Ferguson <jpferg at mit.edu>wrote:

It's more the first step than the last--but yes, you're right. Thanks! --JP On Fri, Feb 6, 2009 at 11:38 PM, Bernard L. Fraga <bfraga at fas.harvard.edu

wrote:

I think it's just that you're missing the last step, which is to associate the draw of "row numbers" with the data for each of those rows. Code like the following might be helpful: yoursample <- samplespace[thedraw,] I haven't checked this, but hopefully this is a start. Good luck! -Bernard ----------------------- Bernard L. Fraga Ph.D. Student, Harvard University Government and Social Policy bfraga at fas.harvard.edu ----------------------- On Feb 6, 2009, at 11:13 PM, John-Paul Ferguson wrote: Hi Everyone, I decided to get an early start on the problem set. I have a question about the sample() command, if anyone has gone that far (this relates to 2.4). In question 2.4, we are told to "Randomly draw 100 datasets with 50 observations each from the data loaded in part 2.2 with replacement...using sample()". I can't figure out how to get sample() to draw *rows* rather than single *values*. That is, if I write samplespace <- matrix(c(y,x1,x2,x3),1000,4) thedraw <- sample(samplespace, 50, replace=TRUE) then I get fifty scalars rather than fifty rows from the matrix samplespace. Can anyone see what I'm getting wrong here? The help on sample() isn't actually much help; it too seems to assume that you'll be drawing numbers. I normally wouldn't post such a "basic" question, but I figure that I can't be the only person who's puzzling out R syntax this week! Best, John-Paul Ferguson PhD Candidate, Economic Sociology MIT Sloan School of Management 50 Memorial Drive, E52-533 Cambridge, MA 02142 617.253.3940 (w) 617.549.8482 (c) http://web.mit.edu/jpferg/www _______________________________________________ gov2001-l mailing list gov2001-l at lists.fas.harvard.edu http://lists.fas.harvard.edu/mailman/listinfo/gov2001-l _______________________________________________ gov2001-l mailing list gov2001-l at lists.fas.harvard.edu http://lists.fas.harvard.edu/mailman/listinfo/gov2001-l

_______________________________________________ gov2001-l mailing list gov2001-l at lists.fas.harvard.edu http://lists.fas.harvard.edu/mailman/listinfo/gov2001-l

-- Diane Kuhn PhD Candidate Department of Government Harvard University dmkuhn at fas.harvard.edu _______________________________________________ gov2001-l mailing list gov2001-l at lists.fas.harvard.edu http://lists.fas.harvard.edu/mailman/listinfo/gov2001-l

Reply

plam＠fas.harvard.edu

4:34 p.m.

Correction: Helpful hints and answers to specific questions (about pieces of the homework perhaps) are encouraged. However, please don't send around whole answers to questions on the homework. On Sat, Feb 7, 2009 at 12:58 PM, Patrick Lam <plam at fas.harvard.edu> wrote:

...

Please refrain from posting answers on the list. Questions are okay though! On Sat, Feb 7, 2009 at 9:08 AM, John-Paul Ferguson <jpferg at mit.edu> wrote:

That is what I ultimately did last night, Diane, with Bernard's idea as the jumping-off point. I'd post the code but I'm not sure whether that's unsporting. JP On Sat, Feb 7, 2009 at 8:42 AM, Diane Kuhn <dmkuhn08 at gmail.com> wrote:

I think the way to think about this might be to draw a random sample of numbers from the numbers of observations in the data (so say to draw a random sample of numbers from the numbers 1-1000), and then once you have your random sample of numbers to call out/subset the dataset with the numbers from your sample: my.sample<-sample(x=1:nrow(data), size=samp.size, replace=TRUE) data.matrix<-data[my.sample,] hope that helps. -diane On Sat, Feb 7, 2009 at 12:58 AM, John-Paul Ferguson <jpferg at mit.edu>wrote:

It's more the first step than the last--but yes, you're right. Thanks! --JP On Fri, Feb 6, 2009 at 11:38 PM, Bernard L. Fraga < bfraga at fas.harvard.edu> wrote: > I think it's just that you're missing the last step, which is to > associate the draw of "row numbers" with the data for each of those rows. > Code like the following might be helpful: > > yoursample <- samplespace[thedraw,] > > I haven't checked this, but hopefully this is a start. Good luck! > > -Bernard > > ----------------------- > Bernard L. Fraga > Ph.D. Student, Harvard University > Government and Social Policy > bfraga at fas.harvard.edu > ----------------------- > > On Feb 6, 2009, at 11:13 PM, John-Paul Ferguson wrote: > > Hi Everyone, > > I decided to get an early start on the problem set. I have a question > about the sample() command, if anyone has gone that far (this relates to > 2.4). > > In question 2.4, we are told to "Randomly draw 100 datasets with 50 > observations each from the data loaded in part 2.2 with replacement...using > sample()". I can't figure out how to get sample() to draw *rows*rather than single > *values*. That is, if I write > > samplespace <- matrix(c(y,x1,x2,x3),1000,4) > thedraw <- sample(samplespace, 50, replace=TRUE) > > then I get fifty scalars rather than fifty rows from the matrix > samplespace. Can anyone see what I'm getting wrong here? The help on > sample() isn't actually much help; it too seems to assume that you'll be > drawing numbers. > > I normally wouldn't post such a "basic" question, but I figure that I > can't be the only person who's puzzling out R syntax this week! > > Best, > John-Paul Ferguson > > PhD Candidate, Economic Sociology > MIT Sloan School of Management > 50 Memorial Drive, E52-533 > Cambridge, MA 02142 > 617.253.3940 (w) > 617.549.8482 (c) > http://web.mit.edu/jpferg/www > _______________________________________________ > gov2001-l mailing list > gov2001-l at lists.fas.harvard.edu > http://lists.fas.harvard.edu/mailman/listinfo/gov2001-l > > > > _______________________________________________ > gov2001-l mailing list > gov2001-l at lists.fas.harvard.edu > http://lists.fas.harvard.edu/mailman/listinfo/gov2001-l > > _______________________________________________ gov2001-l mailing list gov2001-l at lists.fas.harvard.edu http://lists.fas.harvard.edu/mailman/listinfo/gov2001-l

-- Diane Kuhn PhD Candidate Department of Government Harvard University dmkuhn at fas.harvard.edu _______________________________________________ gov2001-l mailing list gov2001-l at lists.fas.harvard.edu http://lists.fas.harvard.edu/mailman/listinfo/gov2001-l

_______________________________________________ gov2001-l mailing list gov2001-l at lists.fas.harvard.edu http://lists.fas.harvard.edu/mailman/listinfo/gov2001-l

-- Patrick Lam Department of Government and Institute for Quantitative Social Science, Harvard University http://www.people.fas.harvard.edu/~plam<http://www.people.fas.harvard.ed…

-- Patrick Lam Department of Government and Institute for Quantitative Social Science, Harvard University http://www.people.fas.harvard.edu/~plam

Reply