Re: st: L-statistic

[n j cox <n.j.cox@durham.ac.uk>]

Joseph Coveney wrote:

> I don't know how much illumination the following provides, but I have the second edition 
> and it has two examples of the L-statistic, one in using "Artificial data-1" (Table 5.10 on 
> Page 162) and the other using "Determination of the ESR" (Table 5.15 on Page 181).  
> Analysis of the latter dataset does produce an anomaly that Nick describes.
< snip >

> --------------------------------------------------------------------------------
>
> clear
> set more off
> tempfile tmp
> insheet using "Determination of the ESR.dat"
> replace v1 = subinstr(subinstr(v1, "  ", " ", .), " ", ",", .)
> outsheet using `tmp', comma nonames noquote
> insheet using `tmp', comma names clear
> erase `tmp'
< snip >

Thanks very much for looking at this so thoroughly.

There is a side issue here which I will comment on briefly.

Joseph's early trickery (code above) reflects the fact that the raw data 
file looks like this:

--------------------------------- start of file
individual fib glob y

1  2.52 38 0
2  2.56 31 0
3  2.19 33 0
4  2.18 31 0
5  3.41 37 0
6  2.46 36 0
7  3.22 38 0
8  2.21 37 0
9  3.15 39 0
<snip>
--------------------------------- end of file

-- in other words, a header and a blank line and a line
for each observation with spaces separating
values. Joseph -insheet-s it, -outfile-s it
as comma-separated and then -infile-s again.

Here is another way to do it. It is no better, but
different.

Copy and paste data into editor:

edit

The result is one string variable, -var1-.

Save the variable names

tokenize `= var1[1]'

and drop the first two lines

drop in 1/2

and then -split- and -destring- in one

split var1, destring

and finish with a renaming

forval i = 1/4 {
	rename var1`i' ``i''
}

and dropping the original:

drop var1

Nick
n.j.cox@durham.ac.uk


*
*   For searches and help try:
*   http://www.stata.com/support/faqs/res/findit.html
*   http://www.stata.com/support/statalist/faq
*   http://www.ats.ucla.edu/stat/stata/