Saturday, 22 December 2007
Thursday, 20 December 2007
This will effect all services we provide, and will seriously restrict what your computers can do.
Stata (32 and 64bit)
RLAB Web sites
I hope the majority of you will be off enjoying you vacation but for those of you who will be working that day please come and see me or the team for advice on what you will be able to do and how you will work.
I apologise for any inconvenience this may cause you.
Monday, 19 November 2007
|Warning||No Send||No Send/|
|Staff and Research Students||250MB||270MB||300MB|
This is an increase of over double the previous limits.
Wednesday, 7 November 2007
1. Use the version command.
If you use this command as standard there won't be a problem of saving to an older version.
2. or use the saveold command. Instead of using, save filename, replace, use saveold filename, replace (this saves the data as stata7, which can be read in 8, 9 and 10).
Thursday, 6 September 2007
We are upgrading our Stata to version 10 and to do this I need to restart the server it is held on. I am going to do this tomorrow morning (Friday 7th Sept.) between 8:30 and 9:00.
Please make sure you quit out of Stata when you leave tonight or it will crash and any data not saved will be lost when I restart the server.
Thanks, and enjoy Stata 10.
Thursday, 16 August 2007
Brendan McConnell Writes:
Hi there computer people,
I’m emailing as I finally figured out a solution to a problem I was having with a dataset, and I thought I’d tell you in case people have the same problem in the future, and I felt like I was banging my head against a brick wall trying to figure it out.
The problem I had was that someone at the DfES messed up a .txt dataset that I was trying to insheet into Stata; instead of the rows being separated by carriage returns, they were separated by tabs (columns were separated by commas) and Stata didn’t like this, as it read it in as a single row. I couldn’t rectify the problem with a “replace all” type command in any word editors as the programs couldn’t load the dataset fully (it about 85 MB, with all the data on a single line). Stat-transfer wouldn’t work either
I finally figured out that I could use the commands hexdump (hexadecimal dump – means nothing to me!) and filefilter from within Stata to alter the dataset before it is opened. It’s so simple, but it worked perfectly for my problem. I’ve attached the relevant bit of my log file for you to see how it works – it just let me replace all tabs with carriage returns, changing the dataset from being 1 line with 87,302,806 characters to 595,377 lines with something like 30 columns. Perfect.
So now I’ve figured it out in Stata, would there have been a better way? Is there a UNIX version of Stat-transfer? Also, is there a command line in the UNIX editor bit (not sure what it’s called) that I could have used to have done the same thing. Irrelevant for this task, but I’d like to know for future reference.
Tanvi Desai Writes:
sounds like you found a very good method.
There are two possible alternatives, but I'm not sure they are an improvement on what you've done. The first method (which I usually use) is to do a find and replace on UNIX...
The other quick and dirty way is to infile into excel first where you can specify your delimiter, and then stat-transfer, but this obviously only works for the smaller datasets because of excel's size limits.
Wednesday, 1 August 2007
From today, 1st August 2007, for testing during August I am setting the main printers to print in EconoMode as standard. This means all printing will use less toner printing out in dark grey as apposed to black. If you need to print out a document that requires the text to be black follow the instructions below.
When the print window comes up select the “properties” button next to the printer.
In the properties window select the “Paper/Quality” tab
Untick the EconoMode option. (See picture below)
Click OK, then print out you document.