[Date Prev][Date Next][Thread Prev][Thread Next][Date index][Thread index]

Re: st: To STATA experts - BOOK

From   Ulrich Kohler <[email protected]>
To   [email protected]
Subject   Re: st: To STATA experts - BOOK
Date   Fri, 14 Sep 2007 11:50:32 +0200

> Please, understand "non-statisticians researchers" as professionals and
> researchers without a background in Math, Statistics, Economy or Computer
> Science. These researchers do not have much time to wait 24h to get an
> answer to their simple {some times idiot] queries. Moreover, I cannot
> spend 1h per day seeking answers and downloading materials on the
> internet. I just wanna a bible "stata for dummies", where I can get an
> answer to [wtf does "floating-point variable" mean?] without needing to
> spend 120 hours of intro to programming.

Ok, if I try to empathize with a Stata newbie, who wants to now more
about floating-point variables, I would probably use one of the
following strategies:

(1) Looking into the subject index of the Stata manuals: 

Unfortunately there is no entry "floating-point variable" in the index.
However there is the entry:

float, [D] data types, [I] data types, [U] 12.2.2 Numeric storage types,
[U] 13.10 Precision and problems therein.

Looking and the first place "[D] data types", I found a rather strange
entry. It says in the very first line "This entry provides a quick
reference  for data types allowed by Stata, See [U] 12 Data for

So, let's look at [U] 12, especially at that parts already mentioned in
the subject index, i.e. section 12.2.2. The first section here is:

"Numbers can be stored in one of five variable types: byte, int, long,
float (the default), or doubles. Bytes are, naturally, stored in 1 byte.
ints are stored in 2 bytes, longs and floats in 4 bytes, and doubles in
8 bytes. The table below shows the minimum and maximum values for each
storage type."

Then there is a table, which does exactely this. 

If I am still unsatisfied with this, I would go to the final entry
mentioned in the subject index: [U] 13.10 Precision and problems
therein. Here I find a lengthy (i.e. 1 page) explanation of what can be
stored in floats, and what not. 

(2) Use Stata's online help 


. search floating point

into Stata reveals as very first entry:

[U]     Chapter
12  . . . . . . . . . . . . . . . . . . . . . . . . . . . Data
        (help datatypes, label, missing, notes, format)

Clicking on "help datatypes" brings up a help file, which somehow 
comprises the section 12.2. and 13.10 of the Users-Guide.

(3) Looking into a Stata-Book

Looking into the subject index of "Kohler/Kreuter, Data Analysis Using
Stata" I find the entry "float" which points to page 100. On page 100 is
the section "Storage types, or, the ghost in the machine". It starts
with the following:

"To conclude this chapter, we should point out a frustrating problem
that you will encounter sooner or later. Before we can explain this
problem, we must make a short technical introduction. Stata
distinguishes alphanumeric variables (strings) from numerical variables
(reals). Strings contain letters and other characters (including
numerals that ar not used as numbers). Reals are numbers. For both types
of variables Stata distinguishes between different storage types ..."

And so on. The section, which is around a length of 1 page, talks a
little bit about storage types and especially about the precision
problem that is also mentioned in entry 13.10 of the Users-Guide. 
In also uses the term "floating point variable" at one place.

(4) Google to "Floating point variable Stata"

Here are the first three links:

The first and third link of these are quite good places to learn
something about floating-point variables. 


I thing I got a good idea of what the term means, and - more importantly - 
what consequences it might have for me as Stata user. Definitely 
I got that idea without needing to "spend 120 hours of intro to programming".

Personally, I liked strategy 3 most. 

My impression is, that the critique is slightly over-the-top. At least
for this example.

True, with all my strategies I did not find anywhere a precise "definition"
of the term floating-point variable. This, btw I can get from Wikipedia. You 
find it below. However, I doubt that this is something the normal Stata
newbie wants to read:

Excerpt from Wikipedia

In computing, floating-point is a numerical-representation system in
which a string of digits (or bits) represents a real number. The most
commonly encountered representation is that defined by the IEEE 754

The name "floating-point" refers to the fact that the radix point
(decimal point, or, more commonly in computers, binary point) can be
placed anywhere relative to the digits within the string. This position
is indicated separately in the internal representation, and
floating-point representation can thus be thought of as a computer
realization of scientific notation.

The advantage of floating-point representation over fixed-point (and
integer) representation is that it supports a much wider range of
values. For example, while a fixed-point representation that allocates
eight decimal digits and two decimal places can represent the numbers
123456.78, 8765.43, 123.00, and so on, a floating-point representation
with eight decimal digits could also represent 1.2345678, 1234567.8,
0.000012345678, 12345678000000000, and so on.

*   For searches and help try:

© Copyright 1996–2024 StataCorp LLC   |   Terms of use   |   Privacy   |   Contact us   |   What's new   |   Site index