11 Feb
2003
11 Feb
'03
2:25 a.m.
all,
I wrote the following function for Q1(a), and got the obviously wrong answer.
any idea as to what I did wrong?
sims<-1000
people<-5
pool<-seq(1,100,1)
same<-0
for(i in 1:sims){
+ comm<-sample(pool, people, replace=FALSE)
+ if(length(unique(comm))<people)
+ same<-same+1}
cat("probability of =5 people being in the same
committee:", same/sims, "\n")
probability of >=5 people being in
the same committee: 0
yongwook
-----------------------------
Yongwook Ryu
PhD Candidate
Department of Government
Harvard University
Tel:617-493-3397
Email: yryu(a)fas.harvard.edu
-----------------------------
11 Feb
11 Feb
7:27 a.m.
Yongwook,
Your sequence "pool" is a list of 100 elements, with values from 1 to 100. I
think you want to change the values in your sequence so that the number of
elements that take on the value "1" corresponds to the number of democrats,
and the rest of the elements (which represent republicans) take on the value
"0" (or some other values that will distinguish between the two).
Your if statement tests if the length of the unique sample is less than 5.
As you have your sampling done now, the sample contains all unique numbers,
so you will never get less than 5 unique numbers from your sample (without
replacement). After you make the changes above you will still want to change
your if statement. You want to record the times that all of the people drawn
from the senate are democrats, or in terms of the above coding convention
(dems = 1), you want to record when your sample contains all 1's.
Hope this help.
John.
On 2/11/03 2:25 AM, "Yongwook Ryu" <yryu(a)fas.harvard.edu> wrote:
all,
I wrote the following function for Q1(a), and got the obviously wrong answer.
any idea as to what I did wrong?
sims<-1000
people<-5
pool<-seq(1,100,1)
same<-0
for(i in 1:sims){
+ comm<-sample(pool, people, replace=FALSE)
+ if(length(unique(comm))<people)
+ same<-same+1}
cat("probability of =5 people being in the
same committee:", same/sims, "\n")
probability of >=5 people
being in the same committee: 0
yongwook
-----------------------------
Yongwook Ryu
PhD Candidate
Department of Government
Harvard University
Tel:617-493-3397
Email: yryu(a)fas.harvard.edu
-----------------------------
_______________________________________________
gov2001-l mailing list
gov2001-l(a)fas.harvard.edu
http://www.fas.harvard.edu/mailman/listinfo/gov2001-l
9:21 a.m.
Yes, you need to come up with an index which separates democrats from
republicans.
Kosuke
On Tue, 11 Feb 2003, John Bright wrote:
Yongwook,
Your sequence "pool" is a list of 100 elements, with values from 1 to 100. I
think you want to change the values in your sequence so that the number of
elements that take on the value "1" corresponds to the number of democrats,
and the rest of the elements (which represent republicans) take on the value
"0" (or some other values that will distinguish between the two).
Your if statement tests if the length of the unique sample is less than 5.
As you have your sampling done now, the sample contains all unique numbers,
so you will never get less than 5 unique numbers from your sample (without
replacement). After you make the changes above you will still want to change
your if statement. You want to record the times that all of the people drawn
from the senate are democrats, or in terms of the above coding convention
(dems = 1), you want to record when your sample contains all 1's.
Hope this help.
John.
On 2/11/03 2:25 AM, "Yongwook Ryu" <yryu(a)fas.harvard.edu> wrote:
all,
I wrote the following function for Q1(a), and got the obviously wrong answer.
any idea as to what I did wrong?
_______________________________________________
gov2001-l mailing list
gov2001-l(a)fas.harvard.edu
http://www.fas.harvard.edu/mailman/listinfo/gov2001-l
sims<-1000
people<-5
pool<-seq(1,100,1)
same<-0
for(i in 1:sims){
+ comm<-sample(pool, people, replace=FALSE)
+ if(length(unique(comm))<people)
+ same<-same+1}
cat("probability of =5 people being in the
same committee:", same/sims, "\n")
probability of >=5 people
being in the same committee: 0
yongwook
-----------------------------
Yongwook Ryu
PhD Candidate
Department of Government
Harvard University
Tel:617-493-3397
Email: yryu(a)fas.harvard.edu
-----------------------------
_______________________________________________
gov2001-l mailing list
gov2001-l(a)fas.harvard.edu
http://www.fas.harvard.edu/mailman/listinfo/gov2001-l
4 p.m.
New subject: [gov2001-l] Q1(a) - "independent" assignment?
Question 1(a) says that "committee assignment is independent" which
means:
Pr(1st selected Democrat) = Pr (2nd selected Democrat | 1st selected
Democrat)
However, these probabilities are not the same:
Pr (1st selected Dem) = 45/100
Pr (2nd selected Democrat | 1st selected Democrat) = 44/99
So, in what sense is the assignment "independent"?
Thanks,
Stan
****************************
Stanislav Markus
Ph.D. Candidate
Harvard University
Department of Government
e: smarkus(a)fas.harvard.edu
t: 617.513.5407
-----Original Message-----
From: gov2001-l-admin(a)fas.harvard.edu
[mailto:gov2001-l-admin@fas.harvard.edu] On Behalf Of Kosuke Imai
Sent: Tuesday, February 11, 2003 9:21 AM
To: John Bright
Cc: Yongwook Ryu; gov2001-l(a)fas.harvard.edu
Subject: Re: [gov2001-l] re: Q1(a)
Yes, you need to come up with an index which separates democrats from
republicans.
Kosuke
On Tue, 11 Feb 2003, John Bright wrote:
_______________________________________________
gov2001-l mailing list
gov2001-l(a)fas.harvard.edu
http://www.fas.harvard.edu/mailman/listinfo/gov2001-l
Yongwook,
Your sequence "pool" is a list of 100 elements, with values from 1 to
100. I
think you want to change the values in your sequence
so that the
number of
elements that take on the value "1"
corresponds to the number of
democrats,
and the rest of the elements (which represent
republicans) take on the
value
"0" (or some other values that will
distinguish between the two).
Your if statement tests if the length of the unique sample is less
than 5.
As you have your sampling done now, the sample
contains all unique
numbers,
so you will never get less than 5 unique numbers from
your sample
(without
replacement). After you make the changes above you
will still want to
change
your if statement. You want to record the times that
all of the people
drawn
from the senate are democrats, or in terms of the
above coding
convention
(dems = 1), you want to record when your sample
contains all 1's.
Hope this help.
John.
On 2/11/03 2:25 AM, "Yongwook Ryu" <yryu(a)fas.harvard.edu> wrote:
> all,
>
> I wrote the following function for Q1(a), and got the obviously
wrong answer.
> any idea as to what I did wrong?
>
>> sims<-1000
>> people<-5
>> pool<-seq(1,100,1)
>> same<-0
>> for(i in 1:sims){
> + comm<-sample(pool, people, replace=FALSE)
> + if(length(unique(comm))<people)
> + same<-same+1}
>> cat("probability of =5 people being in the same committee:",
same/sims, "\n")
probability of
>=5 people being in the same committee: 0
yongwook
-----------------------------
Yongwook Ryu
PhD Candidate
Department of Government
Harvard University
Tel:617-493-3397
Email: yryu(a)fas.harvard.edu
-----------------------------
_______________________________________________
gov2001-l mailing list
gov2001-l(a)fas.harvard.edu
http://www.fas.harvard.edu/mailman/listinfo/gov2001-l
_______________________________________________
gov2001-l mailing list
gov2001-l(a)fas.harvard.edu
http://www.fas.harvard.edu/mailman/listinfo/gov2001-l
4:25 p.m.
New subject: [gov2001-l] Q1(a) - "independent" assignment?
Stan:
I think it is more realistic not to assume independence. No one should be
on the committee twice, right?
Ryan
------------------------------------------
Ryan T. Moore ~ Government & Social Policy
Ph.D. Candidate ~ Harvard University
On Tue, 11 Feb 2003, Stanislav Markus wrote:
Question 1(a) says that "committee assignment is
independent" which
means:
Pr(1st selected Democrat) = Pr (2nd selected Democrat | 1st selected
Democrat)
However, these probabilities are not the same:
Pr (1st selected Dem) = 45/100
Pr (2nd selected Democrat | 1st selected Democrat) = 44/99
So, in what sense is the assignment "independent"?
Thanks,
Stan
****************************
Stanislav Markus
Ph.D. Candidate
Harvard University
Department of Government
e: smarkus(a)fas.harvard.edu
t: 617.513.5407
-----Original Message-----
From: gov2001-l-admin(a)fas.harvard.edu
[mailto:gov2001-l-admin@fas.harvard.edu] On Behalf Of Kosuke Imai
Sent: Tuesday, February 11, 2003 9:21 AM
To: John Bright
Cc: Yongwook Ryu; gov2001-l(a)fas.harvard.edu
Subject: Re: [gov2001-l] re: Q1(a)
Yes, you need to come up with an index which separates democrats from
republicans.
Kosuke
On Tue, 11 Feb 2003, John Bright wrote:
_______________________________________________
gov2001-l mailing list
gov2001-l(a)fas.harvard.edu
http://www.fas.harvard.edu/mailman/listinfo/gov2001-l
_______________________________________________
gov2001-l mailing list
gov2001-l(a)fas.harvard.edu
http://www.fas.harvard.edu/mailman/listinfo/gov2001-l
Yongwook,
Your sequence "pool" is a list of 100 elements, with values from 1 to
100. I
think you want to change the values in your
sequence so that the
number of
elements that take on the value "1"
corresponds to the number of
democrats,
and the rest of the elements (which represent
republicans) take on the
value
"0" (or some other values that will
distinguish between the two).
Your if statement tests if the length of the unique sample is less
than 5.
As you have your sampling done now, the sample
contains all unique
numbers,
so you will never get less than 5 unique numbers
from your sample
(without
replacement). After you make the changes above
you will still want to
change
your if statement. You want to record the times
that all of the people
drawn
from the senate are democrats, or in terms of the
above coding
convention
(dems = 1), you want to record when your sample
contains all 1's.
Hope this help.
John.
On 2/11/03 2:25 AM, "Yongwook Ryu" <yryu(a)fas.harvard.edu> wrote:
> all,
>
> I wrote the following function for Q1(a), and got the obviously
wrong
answer.
> any idea as to what I did wrong?
>
>> sims<-1000
>> people<-5
>> pool<-seq(1,100,1)
>> same<-0
>> for(i in 1:sims){
> + comm<-sample(pool, people, replace=FALSE)
> + if(length(unique(comm))<people)
> + same<-same+1}
>> cat("probability of =5 people being in the same committee:",
same/sims, "\n")
probability of >=5 people being in the same committee: 0
yongwook
-----------------------------
Yongwook Ryu
PhD Candidate
Department of Government
Harvard University
Tel:617-493-3397
Email: yryu(a)fas.harvard.edu
-----------------------------
_______________________________________________
gov2001-l mailing list
gov2001-l(a)fas.harvard.edu
http://www.fas.harvard.edu/mailman/listinfo/gov2001-l
_______________________________________________
gov2001-l mailing list
gov2001-l(a)fas.harvard.edu
http://www.fas.harvard.edu/mailman/listinfo/gov2001-l
4:49 p.m.
New subject: [gov2001-l] Q1(a) - "independent" assignment?
That's right. Nobody should be on the same committee twice.
"independent" here means that the probability of senator i selected into a
committee doesn't depend on that of other senators. In fact, the selection
probability is the same for everyone, hence "uniformly distributed".
Kosuke
On Tue, 11 Feb 2003, Ryan Thomas Moore wrote:
Stan:
I think it is more realistic not to assume independence. No one should be
on the committee twice, right?
Ryan
------------------------------------------
Ryan T. Moore ~ Government & Social Policy
Ph.D. Candidate ~ Harvard University
On Tue, 11 Feb 2003, Stanislav Markus wrote:
Question 1(a) says that "committee
assignment is independent" which
means:
Pr(1st selected Democrat) = Pr (2nd selected Democrat | 1st selected
Democrat)
However, these probabilities are not the same:
Pr (1st selected Dem) = 45/100
Pr (2nd selected Democrat | 1st selected Democrat) = 44/99
So, in what sense is the assignment "independent"?
Thanks,
Stan
****************************
Stanislav Markus
Ph.D. Candidate
Harvard University
Department of Government
e: smarkus(a)fas.harvard.edu
t: 617.513.5407
-----Original Message-----
From: gov2001-l-admin(a)fas.harvard.edu
[mailto:gov2001-l-admin@fas.harvard.edu] On Behalf Of Kosuke Imai
Sent: Tuesday, February 11, 2003 9:21 AM
To: John Bright
Cc: Yongwook Ryu; gov2001-l(a)fas.harvard.edu
Subject: Re: [gov2001-l] re: Q1(a)
Yes, you need to come up with an index which separates democrats from
republicans.
Kosuke
On Tue, 11 Feb 2003, John Bright wrote:
_______________________________________________
gov2001-l mailing list
gov2001-l(a)fas.harvard.edu
http://www.fas.harvard.edu/mailman/listinfo/gov2001-l
_______________________________________________
gov2001-l mailing list
gov2001-l(a)fas.harvard.edu
http://www.fas.harvard.edu/mailman/listinfo/gov2001-l
_______________________________________________
gov2001-l mailing list
gov2001-l(a)fas.harvard.edu
http://www.fas.harvard.edu/mailman/listinfo/gov2001-l
Yongwook,
Your sequence "pool" is a list of 100 elements, with values from 1 to
100. I
think you want to change the values in your
sequence so that the
number of
elements that take on the value "1"
corresponds to the number of
democrats,
and the rest of the elements (which represent
republicans) take on the
value
"0" (or some other values that will
distinguish between the two).
Your if statement tests if the length of the unique sample is less
than 5.
As you have your sampling done now, the sample
contains all unique
numbers,
so you will never get less than 5 unique numbers
from your sample
(without
replacement). After you make the changes above
you will still want to
change
your if statement. You want to record the times
that all of the people
drawn
from the senate are democrats, or in terms of the
above coding
convention
(dems = 1), you want to record when your sample
contains all 1's.
Hope this help.
John.
On 2/11/03 2:25 AM, "Yongwook Ryu" <yryu(a)fas.harvard.edu> wrote:
> all,
>
> I wrote the following function for Q1(a), and got the obviously
wrong
answer.
> any idea as to what I did wrong?
>
>> sims<-1000
>> people<-5
>> pool<-seq(1,100,1)
>> same<-0
>> for(i in 1:sims){
> + comm<-sample(pool, people, replace=FALSE)
> + if(length(unique(comm))<people)
> + same<-same+1}
>> cat("probability of =5 people being in the same committee:",
same/sims, "\n")
probability of >=5 people being in the same committee: 0
yongwook
-----------------------------
Yongwook Ryu
PhD Candidate
Department of Government
Harvard University
Tel:617-493-3397
Email: yryu(a)fas.harvard.edu
-----------------------------
_______________________________________________
gov2001-l mailing list
gov2001-l(a)fas.harvard.edu
http://www.fas.harvard.edu/mailman/listinfo/gov2001-l
_______________________________________________
gov2001-l mailing list
gov2001-l(a)fas.harvard.edu
http://www.fas.harvard.edu/mailman/listinfo/gov2001-l
7:20 p.m.
New subject: [gov2001-l] unbiasedness
How do we go about verifying unbiasedness?
-Ryan
7:32 p.m.
New subject: [gov2001-l] unbiasedness
You know what the true values for beta1, beta2 and sigma^2 are.
So it's unbiased if the simulated values are pretty much right
on the true values.
----- Original Message -----
From: "Ryan Davies" <rdavies(a)fas.harvard.edu>
To: <gov2001-l(a)fas.harvard.edu>
Sent: Tuesday, February 11, 2003 7:20 PM
Subject: [gov2001-l] unbiasedness
How do we go about verifying unbiasedness?
-Ryan
_______________________________________________
gov2001-l mailing list
gov2001-l(a)fas.harvard.edu
http://www.fas.harvard.edu/mailman/listinfo/gov2001-l
7:43 p.m.
New subject: [gov2001-l] unbiasedness
Yes. You check whether the mean of your estimates is equal to the true
value.
Kosuke
On Tue, 11 Feb 2003, Olivia Lau wrote:
You know what the true values for beta1, beta2 and
sigma^2 are.
So it's unbiased if the simulated values are pretty much right
on the true values.
----- Original Message -----
From: "Ryan Davies" <rdavies(a)fas.harvard.edu>
To: <gov2001-l(a)fas.harvard.edu>
Sent: Tuesday, February 11, 2003 7:20 PM
Subject: [gov2001-l] unbiasedness
How do we go about verifying unbiasedness?
-Ryan
_______________________________________________
gov2001-l mailing list
gov2001-l(a)fas.harvard.edu
http://www.fas.harvard.edu/mailman/listinfo/gov2001-l
_______________________________________________
gov2001-l mailing list
gov2001-l(a)fas.harvard.edu
http://www.fas.harvard.edu/mailman/listinfo/gov2001-l
27 Mar
27 Mar
3:50 p.m.
New subject: [gov2001-l] sorting a dataframe
hello, all,
how does one sort a dataframe by values in one column..? sort(), rank(),
and order() seem to work only for vectors..
thanks!
stan
28 Mar
28 Mar
7:56 a.m.
New subject: [gov2001-l] sorting a dataframe
I'm not sure how to massage sort or sort.list to do this on a whole array by
one vector, but ...
If COL is the column that you want to sort on and ARRAY is the array you
want to sort, you can
R> INDEX<-sort.list(COL)
R> SORTED.ARRAY<-ARRAY[INDEX,]
Or in 1 step,
R> SORTED.ARRAY<-ARRAY[sort.list(COL),]
Sort.list(COL) gives the indices for COL in increasing order.
John.
On 3/27/03 3:50 PM, "Stanislav Markus" <smarkus(a)fas.harvard.edu> wrote:
hello, all,
how does one sort a dataframe by values in one column..? sort(), rank(),
and order() seem to work only for vectors..
thanks!
stan
_______________________________________________
gov2001-l mailing list
gov2001-l(a)fas.harvard.edu
http://www.fas.harvard.edu/mailman/listinfo/gov2001-l
7718
days inactive
7763
days old
gov2001@lists.gking.harvard.edu
10 comments
7 participants
participants (7)
-
brightjo＠gse.harvard.edu -
kimai＠fas.harvard.edu -
olau＠fas.harvard.edu -
rdavies＠fas.harvard.edu -
rtmoore＠fas.harvard.edu -
smarkus＠fas.harvard.edu -
yryu＠fas.harvard.edu