Bookmark and Share

Notice: On April 23, 2014, Statalist moved from an email list to a forum, based at

[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index]

Re: st: Stata SE 11.1 memory problem

Subject   Re: st: Stata SE 11.1 memory problem
Date   Sat, 26 Feb 2011 08:12:06 +0200 (EET)

Thank you Eric and Rebecca for your replies. Stata gives me the error
saying that there is no room to add observations. Big dataset is long. I
have checked all you have said. I have enough memory but somehow stata can
not use it. I will ask a tech support. Anyways I have managed to merge the
datasets by dividing the big one into four pieces and then cleaning each
merged piece before merging the next. It is not an ideal solution but
works for now.

 > To add to Eric's comments about operating systems:
> I am using Windows 7 6-bit Professional, and when I first got my
> computer, I couldn't access all of the memory. So, it is still
> possible to not have access to all of the physical memory for your
> computer, even when the operating system is _supposed_ to allow you to
> access it. In my case, it turned out that something had gotten messed
> up in the configuration process at the manufacturer and although I had
> 8GB of memory installed, the Windows OS only recognized 2GB. Some
> setting or something wasn't right; I don't pretend to understand how
> Windows works/does not work with respect to how it recognizes and
> allocates memory. However, tech support on my campus was able to do
> something to fix it, so you might want to check with whomever provides
> tech support for your computer if you find a problem.
> You don't say which OS you're using, but if it's Windows 7, here's how
> you can check the amount of memory the computer recognizes: Control
> panel > System and security > System
> It will tell you how much memory is installed and how much is
> available for program use (i.e. not reserved for the OS). You might
> want to make sure that all the memory you bought is recognized. Also,
> check out the "Memory" tab of the Resource Manager (reached through
> Task Manager) too. It will show you how much memory is in use, how
> much is on "standby" (whatever that means), and how much is "free". In
> my experience, Stata can take up to standby+free. There is a portion I
> can never access and a portion that is marked as "modified". Even if
> I've closed other programs, some of the "modified" remains unavailable
> until I restart. Such is life with Windows.
> As an aside:
> You say your large data set is 244MB and the other is "small",
> presumably less than 244 MB. If that's the case, I don't see why you
> should need to allocate more than 1300 MB. Stata doesn't necessarily
> perform better with more memory
> (
> Regards,
> Rebecca
> On Fri, Feb 25, 2011 at 8:58 PM, Eric Booth <> wrote:
>> <>
>> On Feb 25, 2011, at 6:50 PM, <> wrote:
>>> Thank you for your answer. I have 6GB RAM. I am running nothing else
>>> than
>>> Stata. I have the threads but no help.
>> What type of OS are you running?  One of the threads I mentioned
>> includes some mention about 64-bit OS's that still limit users to 2 or 4
>> GB (e.g., Win7 Starter) (see:
>>   and
>>  ) .
>> Also, you may have not opened any programs, but there are always things
>> running in the background (again, _possibly_ enough to take up 700mb of
>> memory and leave you ~ 1300mb to work with (?) ).   In the first thread
>> link I provided, the Stata technical support response mentions how to
>> look at your task manager (or your OS's equivalent)  to examine your
>> system resources before starting Stata and while trying to set the
>> memory in Stata.
>>> One of the files I have is 244 Mb,
>>> the other one is small. Is there a way to tell to the operating system
>>> to
>>> devote more memory to Stata? I am using very typical financial
>>> databases,
>>> therefore there is probably a way to work with these datasets.
>>  A few years ago, when I was running WinXP (with Stata 9 or 10), I had
>> some luck getting  higher memory allocation by "stepping up" my memory
>> -- try running this code and see if it hits the same 1300m ceiling:
>> ********************!
>> clear
>> forval n = 100(100)6000 {
>>        qui set mem `n'm
>>        qui query mem
>>        di in y  "`r(memory)'"
>> }
>> memory
>> ********************!
>> If you only need a subset of this data you could specify that in your
>> -use- statement (e.g.,  - use var1-var5 using "myfile.dta", clear - ),
>> but that doesn't solve your issue.
>> __
>> Eric A. Booth
>> Public Policy Research Institute
>> Texas A&M University
>> *
>> *   For searches and help try:
>> *
>> *
>> *
> *
> *   For searches and help try:
> *
> *
> *

*   For searches and help try:

© Copyright 1996–2018 StataCorp LLC   |   Terms of use   |   Privacy   |   Contact us   |   Site index