set.seed

Dear All,

I've recently acquired a new MacBook, and while transferring my files and
settings I noticed something I did not expect:

Old Mac:
set.seed(12)
sample(letters[1:10])
[1] "a" "h" "i" "b" "g" "j" "e" "f" "d" "c"

New Mac:
set.seed(12)
sample(letters[1:10])
[1] "b" "g" "c" "f" "e" "i" "d" "j" "h" "a"

For numbers I don't see anything different, the examples from ?set.seed
producing the same results. Both R versions are the same (3.6.2) and both
computers run macOS Catalina 10.15.2.

I would be grateful for any advice,
Adrian

--
Adrian Dusa
University of Bucharest
Romanian Social Data Archive
Soseaua Panduri nr. 90-92
050663 Bucharest sector 5
Romania
https://adriandusa.eu
That's the effect of RNGkind(sample.kind="Rejection") which is the new default. Unfortunately, the older and buggy "Rounding" has a tendency to linger in saved workspaces.

-pd
RNGkind(sample.kind="Rejection")
set.seed(12)
sample(letters[1:10])
[1] "b" "g" "c" "f" "e" "i" "d" "j" "h" "a"
RNGkind(sample.kind="Rounding")
Warning message:
In RNGkind(sample.kind = "Rounding") : non-uniform 'Rounding' sampler used
set.seed(12)
sample(letters[1:10])
[1] "a" "h" "i" "b" "g" "j" "e" "f" "d" "c"
On 23 Jan 2020, at 09:51 , Adrian Du?a via R-SIG-Mac <r-sig-mac at r-project.org> wrote:

Dear All,

I've recently acquired a new MacBook, and while transferring my files and
settings I noticed something I did not expect:

Old Mac:
set.seed(12)
sample(letters[1:10])
[1] "a" "h" "i" "b" "g" "j" "e" "f" "d" "c"

New Mac:
set.seed(12)
sample(letters[1:10])
[1] "b" "g" "c" "f" "e" "i" "d" "j" "h" "a"

For numbers I don't see anything different, the examples from ?set.seed
producing the same results. Both R versions are the same (3.6.2) and both
computers run macOS Catalina 10.15.2.

I would be grateful for any advice,
Adrian

--
Adrian Dusa
University of Bucharest
Romanian Social Data Archive
Soseaua Panduri nr. 90-92
050663 Bucharest sector 5
Romania
https://adriandusa.eu

	[[alternative HTML version deleted]]

_______________________________________________
R-SIG-Mac mailing list
R-SIG-Mac at r-project.org
https://stat.ethz.ch/mailman/listinfo/r-sig-mac

Peter Dalgaard, Professor,
Center for Statistics, Copenhagen Business School
Solbjerg Plads 3, 2000 Frederiksberg, Denmark
Phone: (+45)38153501
Office: A 4.23
Email: pd.mes at cbs.dk  Priv: PDalgd at gmail.com
Dear Peter,

Thanks, it's good to know there is a good explanation for this and also the
sample.kind is now more stable.
Curiously though, I run the very same R version on both computers and I
would have expected RNGkind() to work the same.

Best,
Adrian
That's the effect of RNGkind(sample.kind="Rejection") which is the new
default. Unfortunately, the older and buggy "Rounding" has a tendency to
linger in saved workspaces.
-pd

RNGkind(sample.kind="Rejection")
set.seed(12)
sample(letters[1:10])
 [1] "b" "g" "c" "f" "e" "i" "d" "j" "h" "a"
RNGkind(sample.kind="Rounding")
Warning message:
In RNGkind(sample.kind = "Rounding") : non-uniform 'Rounding' sampler used
set.seed(12)
sample(letters[1:10])
 [1] "a" "h" "i" "b" "g" "j" "e" "f" "d" "c"

On 23 Jan 2020, at 09:51 , Adrian Du?a via R-SIG-Mac <
Dear All,

I've recently acquired a new MacBook, and while transferring my files
and
settings I noticed something I did not expect:

Old Mac:
set.seed(12)
sample(letters[1:10])
[1] "a" "h" "i" "b" "g" "j" "e" "f" "d" "c"

New Mac:
set.seed(12)
sample(letters[1:10])
[1] "b" "g" "c" "f" "e" "i" "d" "j" "h" "a"

For numbers I don't see anything different, the examples from ?set.seed
producing the same results. Both R versions are the same (3.6.2) and
both
computers run macOS Catalina 10.15.2.

I would be grateful for any advice,
Adrian

--
Adrian Dusa
University of Bucharest
Romanian Social Data Archive
Soseaua Panduri nr. 90-92
050663 Bucharest sector 5
Romania
https://adriandusa.eu

      [[alternative HTML version deleted]]

_______________________________________________
R-SIG-Mac mailing list
R-SIG-Mac at r-project.org
https://stat.ethz.ch/mailman/listinfo/r-sig-mac
--
Peter Dalgaard, Professor,
Center for Statistics, Copenhagen Business School
Solbjerg Plads 3, 2000 Frederiksberg, Denmark
Phone: (+45)38153501
Office: A 4.23
Email: pd.mes at cbs.dk  Priv: PDalgd at gmail.com

--
Adrian Dusa
University of Bucharest
Romanian Social Data Archive
Soseaua Panduri nr. 90-92
050663 Bucharest sector 5
Romania
https://adriandusa.eu
On 23 Jan 2020, at 12:15 , Adrian Du?a <dusa.adrian at unibuc.ro> wrote:

Dear Peter,

Thanks, it's good to know there is a good explanation for this and also the sample.kind is now more stable.
Curiously though, I run the very same R version on both computers and I would have expected RNGkind() to work the same.
The most likely explanation is that on one machine you had (like I did)

$ R

R version 3.6.0 (2019-04-26) -- "Planting of a Tree"
Copyright (C) 2019 The R Foundation for Statistical Computing
....blablabla...

[Previously saved workspace restored]
set.seed(12)
sample(letters[1:10])
[1] "a" "h" "i" "b" "g" "j" "e" "f" "d" "c"
RNGkind()
[1] "Mersenne-Twister" "Inversion"        "Rounding"        

...whereas

$ R --vanilla -q
set.seed(12)
sample(letters[1:10])
[1] "b" "g" "c" "f" "e" "i" "d" "j" "h" "a"
RNGkind()
[1] "Mersenne-Twister" "Inversion"        "Rejection"       

-pd
Best,
Adrian

On Thu, Jan 23, 2020 at 11:18 AM peter dalgaard <pdalgd at gmail.com> wrote:
That's the effect of RNGkind(sample.kind="Rejection") which is the new default. Unfortunately, the older and buggy "Rounding" has a tendency to linger in saved workspaces.

-pd

RNGkind(sample.kind="Rejection")
set.seed(12)
sample(letters[1:10])
 [1] "b" "g" "c" "f" "e" "i" "d" "j" "h" "a"
RNGkind(sample.kind="Rounding")
Warning message:
In RNGkind(sample.kind = "Rounding") : non-uniform 'Rounding' sampler used
set.seed(12)
sample(letters[1:10])
 [1] "a" "h" "i" "b" "g" "j" "e" "f" "d" "c"

On 23 Jan 2020, at 09:51 , Adrian Du?a via R-SIG-Mac <r-sig-mac at r-project.org> wrote:

Dear All,

I've recently acquired a new MacBook, and while transferring my files and
settings I noticed something I did not expect:

Old Mac:
set.seed(12)
sample(letters[1:10])
[1] "a" "h" "i" "b" "g" "j" "e" "f" "d" "c"

New Mac:
set.seed(12)
sample(letters[1:10])
[1] "b" "g" "c" "f" "e" "i" "d" "j" "h" "a"

For numbers I don't see anything different, the examples from ?set.seed
producing the same results. Both R versions are the same (3.6.2) and both
computers run macOS Catalina 10.15.2.

I would be grateful for any advice,
Adrian

--
Adrian Dusa
University of Bucharest
Romanian Social Data Archive
Soseaua Panduri nr. 90-92
050663 Bucharest sector 5
Romania
https://adriandusa.eu

      [[alternative HTML version deleted]]

_______________________________________________
R-SIG-Mac mailing list
R-SIG-Mac at r-project.org
https://stat.ethz.ch/mailman/listinfo/r-sig-mac
--
Peter Dalgaard, Professor,
Center for Statistics, Copenhagen Business School
Solbjerg Plads 3, 2000 Frederiksberg, Denmark
Phone: (+45)38153501
Office: A 4.23
Email: pd.mes at cbs.dk  Priv: PDalgd at gmail.com

--
Adrian Dusa
University of Bucharest
Romanian Social Data Archive
Soseaua Panduri nr. 90-92
050663 Bucharest sector 5
Romania
https://adriandusa.eu

Peter Dalgaard, Professor,
Center for Statistics, Copenhagen Business School
Solbjerg Plads 3, 2000 Frederiksberg, Denmark
Phone: (+45)38153501
Office: A 4.23
Email: pd.mes at cbs.dk  Priv: PDalgd at gmail.com

On 23 Jan 2020, at 12:15 , Adrian Du?a <dusa.adrian at unibuc.ro> wrote:
[...]
Curiously though, I run the very same R version on both computers and I
would have expected RNGkind() to work the same.
The most likely explanation is that on one machine you had (like I did)
Yes, that was it.
And easily solved with:

RNGversion(?3.6.2?)

Best,
Adrian

--
Adrian Dusa
University of Bucharest
Romanian Social Data Archive
Soseaua Panduri nr. 90-92
050663 Bucharest sector 5
Romania
https://adriandusa.eu