[James Hassell (StataCorp)]

Thanks for your kind answer! Please correct me if I am mistaken, but that option only works if I run the whole do file, but it doesn't, when I run certain sections by selecting them (which is in 99% the case when I am coding).

Yes, you are correct about how that setting works. Thanks for the clarification. We will look at possibly extending the functionality in the future.

[Clyde Schechter]

As a tangential point to Kazi Bacsi's theme, I really think he should respond to StataCorp's request to contact tech support about these crashes. I have found Stata to be one of the most stable pieces of software around. I can say with completely honesty and certainty that I have not had a single Stata crash since version 12! And I run Stata on 3 different computers (home, office, and a remote server provided by a client I consult to). All of them Windows boxes, but there are even two different versions of Windows among them. Among my Stata-using colleagues the experience has been similar.

[Kazi Bacsi]

As a tangential point to Kazi Bacsi's theme, I really think he should respond to StataCorp's request to contact tech support about these crashes. I have found Stata to be one of the most stable pieces of software around. I can say with completely honesty and certainty that I have not had a single Stata crash since version 12! And I run Stata on 3 different computers (home, office, and a remote server provided by a client I consult to). All of them Windows boxes, but there are even two different versions of Windows among them. Among my Stata-using colleagues the experience has been similar.

Unfortunately I can't reproduce the error, but it happens quite randomly each 2-3 days. The tech support couldn't figure out much from the crash logs and obviously need reproducibility. I am not using any extensions, just the most basic data manipulation commands, it is not related to a single dataset etc., so I'm now quite perplexed.

[Anders Alexandersson]

A discussion of random crashes on Stata 14.2 does not belong to this topic "Stata 15 is here". I suggest Kazi open a new post titled "Random crashes using Stata 14.2 on ..." where ... is the operating system if the problem remains and if tech support stopped their investigation. It could be a hardware problem.

[Fissha Asmare]

I have tried to estimate the mean difference between wage of male and female workers. I have followed two approaches
1. Generate anew variable called wage gap (Male's wage- female's wage) and take the average
2. take the average of male and female separately and take the difference
but the result is not the same. so my concern is; which one is the right result and if not what is the possible problem that the two are not equal.

Thank you

[Marcos Almeida]

Fissha Asmare Welcome to the Statalist/Stata Forum,

It seems your query is not related to this thread.

Being this so, you may wish to start a new topic.

Please read the FAQ, particularly the topic about the best way to post a question so as to entail a helpful reply. Thanks.

[Sean Fiedler]

One question about freduse in Stata 15, I saw a new GUI interface for this command heavily advertised. However after upgrading to Stata 15 I see no interface. Typing

help freduse

doesn't describe any interface, I see nothing in the dropdown menus, and

db freduse

doesn't bring anything up. How do I access this?

Yes, there is new functionality. import fred is a entirely new implementation to access FRED and ALFRED series, and it includes a new and slick graphical interface for searching and browsing for series of interest. It has more new features too.

The old freduse let you specify the series that you wished to load but did not provide ability to search, ability to control aggregation level, a graphical user interface, support for wide versus long formats when importing, selection of date range, and so on. It also used a now-deprecated URL-only way of accessing the series. The old command continues to work for now, but FRED wants developers to use their API for accessing series.

In addition to the new import fred's interface, the new command includes lots of options to control date ranges imported, how series are aggregated, access to historical series, control over wide versus long format, and interacts with FRED via the official FRED API.

The old freduse was not official because it was not up to our standards. It was a placeholder until we could do something better.

[Nick Cox]

File/Import/...Fred works for me. As does

db import fred

[Michael Anbar]

I'm sorry to say that on the face of it, this looks like a very disappointing release.

Disappointing indeed. The complete lack of "modern" debugging tools in Stata (I wouldn't consider breakpoints modern, since they've been around since the 1960s, but I digress) is a major drawback. -set trace on- and -pause- are inferior substitutes, to say the least. Even basic features like autocomplete in the do-file editor are still missing from Stata, years after even some of the most basic text editors, not to mention IDEs, have implemented such features.

[Joseph Coveney]

Debugger features would be good, especially if you do a lot with Mata.

But when I scan through the New in Stata 15 page, I'm glad that StataCorp invested the limited resources that it has as it did. I'll take menl over autocomplete any day of the week.

[Emilio Matuk]

The limit on the size of the weighting matrices in the Sp commands is the limit of matrices in Mata -- which is 2^48 - 1. Obviously, you will reach memory limits on your computer before you make a square matrix of that size. The limit of 11,000 is max_matsize, which only applies to matrices created in Stata, not those created in Mata.

So the effective limit on the size of weighting matrices is the RAM of your computer.

I'm working now in Stata 15. I was able to create a square matrix with dimension 21,689 after day and a half (see command below)

*** spmatrix create idistance W, normalize(minmax)

Later i was able to run a regression with heterocedasticity corrected (see command below)

*** spregress WAGE FR_1-FR_3 PER WAGE_M, gs2sls heteroskedastic

But when I tried dependent variable lagged option (see command below) , I crashed due to memory limitations.

*** spregress WAGE FR_1-FR_3 PER WAGE_M, gs2sls dvarlag(W) heteroskedastic

What suprises me is why I was able to run SPMATRIX but I'mm not able to run SPREGRESS with DVARLAG option.

[Emilio Matuk]

I have tried to estimate the mean difference between wage of male and female workers. I have followed two approaches
1. Generate anew variable called wage gap (Male's wage- female's wage) and take the average
2. take the average of male and female separately and take the difference
but the result is not the same. so my concern is; which one is the right result and if not what is the possible problem that the two are not equal.

Thank you

I will not suggest to do average in any circumstance. Better to work with the full data set and try to evaluate intercept dummy and slope dummy

[Ayesha Sajjad]

Hi all,
I opened my stata 15 today and the simple mi command gives the following error:
mi impute chained (regress) var1 var2 (logit) var3 (ologit) var4 , add(10)

invalid freeparm argument;
only "on" or "off" is allowed
error occurred during imputation of var1 var2 var3 var4 on m = 1

A little help with this would be appreciated.

[Nick Cox]

postdoc nl: Please use your full real name here as requested. https://www.statalist.org/forums/help#realnames

[Clyde Schechter]

I would also add that this thread is starting to get a little diffuse. Some of the more recent posts are really requests for help with specific issues arising from specific commands, rather than being commentary on Stata 15 as a whole, what is liked or disliked about it, how it compares to earlier versions or competing statistical packages etc. Posts that deal with specific problems running specific commands would be better posted as separate threads, in my opinion. If nothing else, it will make it much easier for people searching for those particular problems to find them.