Google Groups no longer supports new Usenet posts or subscriptions. Historical content remains viewable.
Dismiss

newbie: solving variable in exponent of a scientific notated term

0 views
Skip to first unread message

Mike

unread,
Jan 14, 2007, 12:51:26 AM1/14/07
to
Hi All,

I can't seem to solve the following equation on paper without the help of
Matlab ... but I want to understand where I am going wrong here when
attempting to solve for Y:

359,259 = 19.97 x 10^ (0.1977(Y-1960))


My approach is to take the log of both sides:

log(359,259) = log ( 19.97 x 10^ (0.1977(Y-1960)) )

log(359,259) = (0.1977(Y-1960)) * log (19.97x10)

(0.1977(Y-1960)) = log(359,259) / log (19.97x10)

Y-1960 = log(359,259) / ( log (19.97x10) * 0.1977)

Y = log(359,259) / ( log (19.97x10) * 0.1977) + 1960

Y = 1972 which is incorrect. The answer should be 1981.5.

I suspect when I take the log of term in scientific notation which has Y in
the exponent. Can someone please clarify how you properly take the log of
scientific notated term? I did find that if I set the right hand log to:
log(19.97), I got approx the right answer.

However, when I applied the same "logic" to another similar equation:

65705 = 1027 x 10^(0.1505(Y-1970))

it didn't work so I suspect, there's more too it. When I do the same with
an exponential term: e^(x+5) then my resulting log term is: (x+5)*log(e)
and this is correct so it really must have something to do with the way I am
taking of the log of a scientifically notated term.

Thanks in advance, Mike.


Christopher Creutzig

unread,
Jan 14, 2007, 5:40:45 AM1/14/07
to
Mike wrote:

> log(359,259) = log ( 19.97 x 10^ (0.1977(Y-1960)) )
>
> log(359,259) = (0.1977(Y-1960)) * log (19.97x10)

No, check your parentheses and operator precedences.

log(19.97 x 10^(0.1977*(Y-1960))
= log(19.97) + 0.1977*(Y-1960)*log(10)


Regards,
Christopher

Jean-Marc Gulliet

unread,
Jan 14, 2007, 6:19:38 AM1/14/07
to

Mike,

Your approach is correct. However, you seem to have forgotten the
product rule for logarithms: if a and b are two reals such that a*b>0,
then log(a*b) == log(a) + log(b). In other words, the logarithm of a
product is equal to the sum of the logarithms of the terms.

HTH,
Jean-Marc

Christopher Creutzig

unread,
Jan 14, 2007, 10:50:53 AM1/14/07
to
Jean-Marc Gulliet wrote:

> Your approach is correct. However, you seem to have forgotten the
> product rule for logarithms: if a and b are two reals such that a*b>0,
> then log(a*b) == log(a) + log(b). In other words, the logarithm of a

Minor correction: The condition should read “if a > 0 or b > 0.”
0=log(1)=log(-1*-1)=/=log(-1)+log(-1)=2*I*PI. (I regard log as a
single-valued, well-defined function on C\{0}. If you don't, the
equation does not need a restriction.)


Regards,
Christopher

Christopher Creutzig

unread,
Jan 14, 2007, 11:16:21 AM1/14/07
to
Christopher Creutzig wrote:

> Minor correction: The condition should read “if a > 0 or b > 0.”

Or, to give a condition that is both necessary and sufficient:
-PI<arg(a)+arg(b)<=PI.


Regards,
Christopher

Mike

unread,
Jan 18, 2007, 11:32:36 AM1/18/07
to
Thank you everyone !

"Christopher Creutzig" <chris...@creutzig.de> wrote in message
news:45aa5755$0$18834$9b4e...@newsspool4.arcor-online.net...

0 new messages