[Bjarte Aagnes]

doedit https://gist.githubusercontent.com/...

like

do https://gist.githubusercontent.com/...

[Rich Goldstein]

it would be great if -mkspline- (both the linear splines and the restricted cubic splines) could be integrated into the -margins- system

[FernandoRios]

Hi Rich Goldstein
did you consider "f_able" (ssc). It comes with my own versions of mkspline so margins can recognize the created variables.
Here the PPT
https://www.stata.com/meeting/us20/s...ios-Avila1.pdf

[Rich Goldstein]

Hi FernandoRios - I will check it out

[rollanders]

I would like to be able to zoom in on the Viewer or to change font size. It is sometimes hard to read on high-resolution screens. I also want line breaks when shrinking the Viewer window. I tried to solve the problem with resolution by printing the file to pdf, but then came the problem with the long lines that were cut and not shown.

[William Lisowski]

#440 -

In Stata 17 for Mac, in the menu Stata > Preferences > Windows > Viewer the default size of the font can be changed.

Again in Stata 17 for Mac, after changing the Viewer window size, in the menu View > Viewer the Viewer window can be reloaded, which causes the lines to re-wrap to the new window size.

macOS handles zooming at the operating system level in System Preferences > Trackpad by enabling a two-finger pinch gesture to zoom the screen.

[Manuel Ferraro]

Manuel Ferraro Perhaps you may want to look into "f_able". I wrote this utility to correctly handle alternative transformations (beyond polynomials) to correctly estimate marginal effects.
here the paper explaining the idea: https://www.stata-journal.com/articl...article=st0628
and the command can be downloaded from ssc

FernandoRios Thank you for your feedback. Could you please have a look at my post and check whether "f_able" would be helpful in that situation?

https://www.statalist.org/forums/for...dds-for-spline

[Anne Tremblay]

Hi, It would be extremely useful if we could put a password in Stata to protect our databases. Please add this option.

[William Lisowski]

#443 -

An alternative is to use an encrypted flash drive or encrypted disk image to store your sensitive data. That way you give the password one time when the drive is mounted, and for the rest of your session the operating system handles decryption as the files are read and encryption as new files are created, and when the disk is unmounted, the data is protected by the password. And with a flash drive, when you've removed it from your system, it can be stored in a more secure location if need be.

[Jonathan Batty]

As I make more use of frames, it would be helpful to have a pane to display these (and e.g. dimensions, size in memory, saved/unsaved status). It would be great if this could be docked to one side of the main screen, analogous to the variables pane.
I don't personally find the Frames Manager (Data > Frames Manager) that useful. Thanks!

[John Mullahy]

Might there be a way for –bootstrap– and –jackknife– to provide better (=more descriptive) names for the variables they create using the saving(...) option when those variables correspond to parameters associated with factor variables?

sysuse auto
tempfile tjk tbs

qui bootstrap, saving(`tbs', replace): reg price mpg i.foreign
use `tbs'
sum
sysuse auto
qui bootstrap, saving(`tbs', replace): reg price mpg foreign
use `tbs'
sum

sysuse auto
qui jackknife, saving(`tjk', replace): reg price mpg i.foreign
use `tjk'
sum
sysuse auto
qui jackknife, saving(`tjk', replace): reg price mpg foreign
use `tjk'
sum

. sysuse auto
(1978 automobile data)

. tempfile tjk tbs

.
. qui bootstrap, saving(`tbs', replace): reg price mpg i.foreign

. use `tbs'
(bootstrap: regress)

. sum

    Variable |        Obs        Mean    Std. dev.       Min        Max
-------------+---------------------------------------------------------
      _b_mpg |         50    -305.208    68.45507  -469.7973  -139.6592
       _bs_2 |         50           0           0          0          0
       _bs_3 |         50    1836.401    600.9441   406.5179   3537.573
     _b_cons |         50    12093.28    1641.614   7933.012   15650.12

. sysuse auto
(1978 automobile data)

. qui bootstrap, saving(`tbs', replace): reg price mpg foreign

. use `tbs'
(bootstrap: regress)

. sum

    Variable |        Obs        Mean    Std. dev.       Min        Max
-------------+---------------------------------------------------------
      _b_mpg |         50   -297.9724     67.2555  -444.5148  -157.0371
  _b_foreign |         50    1775.099    577.1984   110.2294   2845.707
     _b_cons |         50       11983    1453.518   8555.752   14720.93

.
. sysuse auto
(1978 automobile data)

. qui jackknife, saving(`tjk', replace): reg price mpg i.foreign

. use `tjk'
(jackknife: regress)

. sum

    Variable |        Obs        Mean    Std. dev.       Min        Max
-------------+---------------------------------------------------------
      _b_mpg |         74   -294.2683    7.639854  -330.1661  -273.4005
       _jk_2 |         74           0           0          0          0
       _jk_3 |         74    1767.329    74.31322   1494.767   1996.184
     _b_cons |         74    11906.63    169.9526   11382.11    12618.6

. sysuse auto
(1978 automobile data)

. qui jackknife, saving(`tjk', replace): reg price mpg foreign

. use `tjk'
(jackknife: regress)

. sum

    Variable |        Obs        Mean    Std. dev.       Min        Max
-------------+---------------------------------------------------------
      _b_mpg |         74   -294.2683    7.639854  -330.1661  -273.4005
  _b_foreign |         74    1767.329    74.31322   1494.767   1996.184
     _b_cons |         74    11906.63    169.9526   11382.11    12618.6

.
.
.
.
end of do-file

[David Tsu]

Can we have a command that can export multiple figures and tables into a single word/excel/latex file and is as flexible as outreg2?

[Niels Henrik Bruun]

So far as I can see, this has not been suggested.
It would be immensely nice if -import sas in n1/n2 using my_sas_dataset- worked faster. It seems like the whole dataset is read before the "in n1/n2" selection is performed. Possibly likewise for -import spss-.

[Niels Henrik Bruun]

Regarding -merge-, it would be nice to merge two datasets on a set of key variables without the keys having the same name. The syntax could be like -merge 1:1 master_key1=using_key1 [master_key2=using_key2]- with an option for keeping the using keys (default is to keep the master keys)

[Bjarte Aagnes]

1. value labels for integers > c(maxlong)
2. bigint datatype (ref #407)
3. value labels for str vars