scientific notation

[bmnice]

Kindly advise the best way to write grams per 0.5 m2

  thanx


  ===================================================================
  "Jesus Christ is the SAME...yesterday, today, and forever!"  Hebrews 13:8
Mr B Moyo
  Fixed  [phone]
  Mobile [phone]
  Fax     [phone]
  ===================================================================



---------------------------------
No need to miss a message. Get email on-the-go
with Yahoo! Mail for Mobile. Get started.

[Spousta Jan]

Multiply it by two and write it as grams per m2.

:-)

Jan

-----Original Message-----
From: SPSSX(r) Discussion [mailto:[hidden email]] On Behalf Of
Bethwell Moyo
Sent: Monday, February 05, 2007 4:36 PM
To: [hidden email]
Subject: scientific notation

Kindly advise the best way to write grams per 0.5 m2

  thanx


  ===================================================================
  "Jesus Christ is the SAME...yesterday, today, and forever!"  Hebrews
13:8 Mr B Moyo
  Fixed  [phone]
  Mobile [phone]
  Fax     [phone]
  ===================================================================



---------------------------------
No need to miss a message. Get email on-the-go with Yahoo! Mail for
Mobile. Get started.

[Richard Ristow]

At 10:36 AM 2/5/2007, Bethwell Moyo wrote:

>Kindly advise the best way to write grams per 0.5 m2

About the only way (though I like Jan's riposte) is to create a
variable in which you put the number, stripped of its units; and give
the units in the variable label. I'm making this up, but something like

VAR LABEL
   ConsE46 'Palladium (gm/0.5 m2)'.

Broadly, most values we analyze in SPSS come from measurements or
counts, and most inherently have a description (what they represent)
and 'dimensions', or units:

"Annual income, $1000"
"Yield of corn, bushels/acre"

It would be helpful if SPSS variables could be dimensional quantities,
i.e. have units as attributes, rather than be treated as pure numbers.
With SPSS 14 and 15, where user-defined variable attributes can be
added to the data dictionary, one option is to add such an attribute
and put the units in it. I haven't explored this; I'm not sure how far
anyone has.

(I think that, so far, user-defined attributes don't display nearly as
automatically with the variable as variable labels do, which is an
argument for including the units in the variable label.)
..........
To rhapsodize:

Philosophically, the units should be carried forward in computations.
For example, if you have (I'm making up some of the syntax),

VAR LABEL
    YIELD  "Yield of corn"
    APPL_N "Nitrogen fertilizer applied".
VAR UNITS
    YIELD  "Bushels/acre"
    APPL_N "Pounds per acre".

then if you regress YIELD against APPL_N, with or without other
independent variables, the regression coefficient has units

"(Bushels/acre)/(Pounds/acre)"

But it's easier to say that this is desirable, than it is to develop a
convention for doing it. (Would it have to recognize that "hectare" and
"acre" are different units for the same thing, and recognize they can
be added? Should it cancel "/acre" from the numerator and the
denominator, above?)

End of rhapsody. But analysts should remember that regression
coefficients (as one example) DO have units, if the variables do, and
interpret them accordingly.

[Peck, Jon]

Strictly as a formatting convention, you can use the Custom Currency formats for this.  You can define a CC format in Edit/Options or with a SET command and provide an arbitrary prefix and/or suffix.  There are five formats available for definition: CCA ... CCE.  Note, however, that while the format is saved with the data (the variable has format CCA, for example), the actual definition of the format is a preference setting and belongs to the user, not the data.

HTH,
Jon Peck

-----Original Message-----
From: Richard Ristow [mailto:[hidden email]]
Sent: Monday, February 05, 2007 11:05 AM
To: Bethwell Moyo; [hidden email]
Cc: Spousta Jan; Peck, Jon
Subject: Re: scientific notation

At 10:36 AM 2/5/2007, Bethwell Moyo wrote:

>Kindly advise the best way to write grams per 0.5 m2

About the only way (though I like Jan's riposte) is to create a
variable in which you put the number, stripped of its units; and give
the units in the variable label. I'm making this up, but something like

VAR LABEL
   ConsE46 'Palladium (gm/0.5 m2)'.

Broadly, most values we analyze in SPSS come from measurements or
counts, and most inherently have a description (what they represent)
and 'dimensions', or units:

"Annual income, $1000"
"Yield of corn, bushels/acre"

It would be helpful if SPSS variables could be dimensional quantities,
i.e. have units as attributes, rather than be treated as pure numbers.
With SPSS 14 and 15, where user-defined variable attributes can be
added to the data dictionary, one option is to add such an attribute
and put the units in it. I haven't explored this; I'm not sure how far
anyone has.

(I think that, so far, user-defined attributes don't display nearly as
automatically with the variable as variable labels do, which is an
argument for including the units in the variable label.)
..........
To rhapsodize:

Philosophically, the units should be carried forward in computations.
For example, if you have (I'm making up some of the syntax),

VAR LABEL
    YIELD  "Yield of corn"
    APPL_N "Nitrogen fertilizer applied".
VAR UNITS
    YIELD  "Bushels/acre"
    APPL_N "Pounds per acre".

then if you regress YIELD against APPL_N, with or without other
independent variables, the regression coefficient has units

"(Bushels/acre)/(Pounds/acre)"

But it's easier to say that this is desirable, than it is to develop a
convention for doing it. (Would it have to recognize that "hectare" and
"acre" are different units for the same thing, and recognize they can
be added? Should it cancel "/acre" from the numerator and the
denominator, above?)

End of rhapsody. But analysts should remember that regression
coefficients (as one example) DO have units, if the variables do, and
interpret them accordingly.