J
john sumner
And you are nothing but a google troll whomis going into my bozo bin *PLONK*
And you are nothing but a google troll whomis going into my bozo bin *PLONK*
H
Harlan Grove
(e-mail address removed) wrote...
....
They can sum & they can count. Why, they can even average!
Sarcasm aside, any software that can be automated and can string
together arithmetic operations can perform general numerical analysis.
It's not so much the operators and functions that distinguish databases
and spreadsheets, it's the data referencing. There are some tasks for
which RDMBS data structure is beneficial, but there are many
calculation tasks that require plain multidimensional arrays. Sometimes
the only relationship between the data elements in arrays is just index
order.
While databases do many things well, they don't do such things as
matrix arithmetic or optimization well. Those are tasks which require
iterative procedures on simple matrices or indexed arrays. There's a
considerable amount of statistical, engineering and econometric
analysis that involves matrix and array processing. Certainly RDBMS's
can provide the data storage facility, but you're clearly outside your
ken if you believe SQL queries can produce partial correlation
coefficients from stepwise regression.
Yes, there are add-on tools, many of which are proprietary, most of
which cost extra, and nearly all of which aren't found on average Excel
users' PCs. Besides, the choice of add-on product depends on what the
user prefers to use based on a number of different criteria, including
their familiarity with the add-on products available, but usually not
with much weight given to Aaron's Orthodoxy.
OK, I'll make this simple for you. Here's a table representing a 4x4
matrix. Show us, Oh Great Database Sage, the DBMS steps you'd take to
calculate its determinant and inverse. [Treat sequences of spaces and
underscores as field separators, but ignore the leading ones.]
Fld1__Fld2__Fld3__Fld4
-7__ -24___ -59___ 58
-78__ 65___ 56___ 22
59__ -60___ -31___ 47
35__ -21___ 10____ 1
Something tells me 99.999% of *rational* Access users would perform
this task by creating an Excel application instance and using its
WorksheetFunction.MInverse and WorksheetFunction.MDeterm to perform
these calculations, but I'm asking you to see how one irrational Access
user would do it.
Dunno about the other respondent, but I am. See the challenge above.
See my previous challenge to which you never responded (link below).
For a change of pace, quit talking out your backside and PUT UP OR SHUT
UP.
http://groups.google.com/group/microsoft.public.excel/msg/2c1e73ffd974b4d2?dmode=source&hl=en
....
They can sum & they can count. Why, they can even average!
Sarcasm aside, any software that can be automated and can string
together arithmetic operations can perform general numerical analysis.
It's not so much the operators and functions that distinguish databases
and spreadsheets, it's the data referencing. There are some tasks for
which RDMBS data structure is beneficial, but there are many
calculation tasks that require plain multidimensional arrays. Sometimes
the only relationship between the data elements in arrays is just index
order.
While databases do many things well, they don't do such things as
matrix arithmetic or optimization well. Those are tasks which require
iterative procedures on simple matrices or indexed arrays. There's a
considerable amount of statistical, engineering and econometric
analysis that involves matrix and array processing. Certainly RDBMS's
can provide the data storage facility, but you're clearly outside your
ken if you believe SQL queries can produce partial correlation
coefficients from stepwise regression.
Yes, there are add-on tools, many of which are proprietary, most of
which cost extra, and nearly all of which aren't found on average Excel
users' PCs. Besides, the choice of add-on product depends on what the
user prefers to use based on a number of different criteria, including
their familiarity with the add-on products available, but usually not
with much weight given to Aaron's Orthodoxy.
OK, I'll make this simple for you. Here's a table representing a 4x4
matrix. Show us, Oh Great Database Sage, the DBMS steps you'd take to
calculate its determinant and inverse. [Treat sequences of spaces and
underscores as field separators, but ignore the leading ones.]
Fld1__Fld2__Fld3__Fld4
-7__ -24___ -59___ 58
-78__ 65___ 56___ 22
59__ -60___ -31___ 47
35__ -21___ 10____ 1
Something tells me 99.999% of *rational* Access users would perform
this task by creating an Excel application instance and using its
WorksheetFunction.MInverse and WorksheetFunction.MDeterm to perform
these calculations, but I'm asking you to see how one irrational Access
user would do it.
....
Dunno about the other respondent, but I am. See the challenge above.
See my previous challenge to which you never responded (link below).
For a change of pace, quit talking out your backside and PUT UP OR SHUT
UP.
http://groups.google.com/group/microsoft.public.excel/msg/2c1e73ffd974b4d2?dmode=source&hl=en
H
Harlan Grove
(e-mail address removed) wrote...
....
But the number of users seems to be your primary reason for claiming
VB/VBA/VBS is the best programming language. Why not try some logical
consistency for a change?
Obviously. If most people are writing code in javascript, then the most
legacy code would be in javascript, no? Most code maintenance, in the
short term at least, would require some familiarity with javascript,
no?
Then again, why expect rationality from you? You'd no doubt advocate
ignoring any prior investment in code in other languages, so write all
apps from scratch in VBS.
This doesn't mean that javascript (or COBOL 30 years ago) was the best
language. Unlike you, I haven't made the argument that X is good
because so many people use it. It does seem many people find BASIC's
verbosity comforting relative to the terseness of C and its
descendants. Myself, I use BASIC only because there's no alternative in
some cases.
If only you could demonstrate that you know how to do any number
crunching.
How would you identify statistically significant seasonality in sales
data for a given product?
....
But the number of users seems to be your primary reason for claiming
VB/VBA/VBS is the best programming language. Why not try some logical
consistency for a change?
Obviously. If most people are writing code in javascript, then the most
legacy code would be in javascript, no? Most code maintenance, in the
short term at least, would require some familiarity with javascript,
no?
Then again, why expect rationality from you? You'd no doubt advocate
ignoring any prior investment in code in other languages, so write all
apps from scratch in VBS.
This doesn't mean that javascript (or COBOL 30 years ago) was the best
language. Unlike you, I haven't made the argument that X is good
because so many people use it. It does seem many people find BASIC's
verbosity comforting relative to the terseness of C and its
descendants. Myself, I use BASIC only because there's no alternative in
some cases.
If only you could demonstrate that you know how to do any number
crunching.
How would you identify statistically significant seasonality in sales
data for a given product?
J
Jay Petrulis
Harlan said:
--- snip a bunch of stuff ---
Aaron,
If you can, please show how you would do Harlan's latest challenge in
Access. Please do not resort to some generality like you would use MDX
through Access. Stick to Access and give step by step details. No
broad overview, provide a specific answer.
Other posters who claim as much as you do are held to the same
standards. Just saying you can do this or do that does not make it so.
Please prove your competence with data analysis in Access, which
*you* claim.
Ignore the Excel angle. Please do not even consider it in your reply.
I would like to know how to do it in Access alone.
I do not care how long you have been working on whatever projects and
you can dispense with the preamble rants about anything and everything
Excel (Do you think it is a disease?) You need not worry if I produce
the same report every week. Just answer the challenge from Harlan.
Thanks,
Jay
T
Tony Toews
Aaron
If you could respond in a decent fashion without flipping your lid you'd get better
responses.
Tony
--
Tony Toews, Microsoft Access MVP
Please respond only in the newsgroups so that others can
read the entire thread of messages.
Microsoft Access Links, Hints, Tips & Accounting Systems at
http://www.granite.ab.ca/accsmstr.htm
A
aaron.kempf
im sorry i dont 'respect' your investment in existing languages.
you are the asshole that is preaching the end of vb-- you started this
arg wimpo
there is no reason for anyone to learn java for any reason. I mean--
clientside is a PITA and it can't compete with good server-side code.
statisitically significant seasonality.. HMMM... well i'd take the MSDE
that ships with ACCESS (or is free from www.asp.net) and i'd use the
stdev function
you are the asshole that is preaching the end of vb-- you started this
arg wimpo
there is no reason for anyone to learn java for any reason. I mean--
clientside is a PITA and it can't compete with good server-side code.
statisitically significant seasonality.. HMMM... well i'd take the MSDE
that ships with ACCESS (or is free from www.asp.net) and i'd use the
stdev function
A
aaron.kempf
http://www.ma.utexas.edu/mpej/Vol/1/6.supp/Section.6/linear/minverse
i'd translate this from java into vb or TSQL. and then use it in a
database? i'd keep the function in one place and it would be hella easy
to use.
technically how would i do this?
i'd take my handy-dandy MSDE install (free with windows basically) and
make a UDF that did this.
then i could take my precious little function and use it whenever i
wanted; and my friends could share it with me.
you see-- databases are designed for multiple users.
your gay-ass spreadsheet program can't even handle a single power user.
procedure minverse(var m1,m2: matrix);
var
i,j,k: integer;
s1,s2,s3: scalar;
m0: matrix;
begin { minverse }
write('(minverse'); flush(output);
m0:=m1;
for i:=0 to lmax do for j:=0 to lmax do szero(m2[i,j]);
for i:=0 to lmax do m2[i,i]:=sone;
for i:=0 to lmax do begin { simple Gauss-Jordan }
s1:=m0[i,i];
m0[i,i]:=sone;
for j:=i+1 to lmax do begin
squot(m0[i,j],s1,s2);
m0[i,j]:=s2;
end;
for j:=0 to lmax do begin
squot(m2[i,j],s1,s2);
m2[i,j]:=s2;
end;
for k:=0 to lmax do if k<>i then begin
s1:=m0[k,i];
szero(m0[k,i]);
for j:=i+1 to lmax do begin
sprod(s1,m0[i,j],s2);
sdiff(m0[k,j],s2,s3);
m0[k,j]:=s3;
end;
for j:=0 to lmax do begin
sprod(s1,m2[i,j],s2);
sdiff(m2[k,j],s2,s3);
m2[k,j]:=s3;
end;
end;
end;
write(')'); flush(output);
end { minverse };
i'd translate this from java into vb or TSQL. and then use it in a
database? i'd keep the function in one place and it would be hella easy
to use.
technically how would i do this?
i'd take my handy-dandy MSDE install (free with windows basically) and
make a UDF that did this.
then i could take my precious little function and use it whenever i
wanted; and my friends could share it with me.
you see-- databases are designed for multiple users.
your gay-ass spreadsheet program can't even handle a single power user.
procedure minverse(var m1,m2: matrix);
var
i,j,k: integer;
s1,s2,s3: scalar;
m0: matrix;
begin { minverse }
write('(minverse'); flush(output);
m0:=m1;
for i:=0 to lmax do for j:=0 to lmax do szero(m2[i,j]);
for i:=0 to lmax do m2[i,i]:=sone;
for i:=0 to lmax do begin { simple Gauss-Jordan }
s1:=m0[i,i];
m0[i,i]:=sone;
for j:=i+1 to lmax do begin
squot(m0[i,j],s1,s2);
m0[i,j]:=s2;
end;
for j:=0 to lmax do begin
squot(m2[i,j],s1,s2);
m2[i,j]:=s2;
end;
for k:=0 to lmax do if k<>i then begin
s1:=m0[k,i];
szero(m0[k,i]);
for j:=i+1 to lmax do begin
sprod(s1,m0[i,j],s2);
sdiff(m0[k,j],s2,s3);
m0[k,j]:=s3;
end;
for j:=0 to lmax do begin
sprod(s1,m2[i,j],s2);
sdiff(m2[k,j],s2,s3);
m2[k,j]:=s3;
end;
end;
end;
write(')'); flush(output);
end { minverse };
A
aaron.kempf
and you see.. cobol was know by what 100k people?
VB is the most popular language in the world-- for good reason. because
of spreadsheet dorks like you.
your problem harlan is that you just dont have any comprehension that
a) i can do any math you can but faster
b) im sick and tired of cutting and pasting the same formula 100 times
per day
c) it's less efficient to have 200 copies of the same formula
1) sheer storage
2) accuracy
3) development of these 'super-complex spreadsheets (gag)'
i just think that you guys are a bunch of fucking idiots.
grow up and lose the training wheels.
START BUILDING _SOLUTIONS_ instead of building the same damn XLS week
in and week out
VB is the most popular language in the world-- for good reason. because
of spreadsheet dorks like you.
your problem harlan is that you just dont have any comprehension that
a) i can do any math you can but faster
b) im sick and tired of cutting and pasting the same formula 100 times
per day
c) it's less efficient to have 200 copies of the same formula
1) sheer storage
2) accuracy
3) development of these 'super-complex spreadsheets (gag)'
i just think that you guys are a bunch of fucking idiots.
grow up and lose the training wheels.
START BUILDING _SOLUTIONS_ instead of building the same damn XLS week
in and week out
H
Harlan Grove
(e-mail address removed) wrote...
Semantics. No question that they *SHOULD* fix bugs. Lord knows there
are lots of very longstanding bugs in Excel. Problem may be that these
bugs have been around so long that the Excel programmers may consider
them 'features' at this point.
But the only thing Microsoft *NEEDS* to do is earn money. If they can
rake in revenues without fixing bugs, why would they bother? One thing
that must be recognized about Microsoft: no matter how nice & upright
some Microsoft employees may be, Microsoft's corporate culture is, to
be charitable, amoral. Microsoft as an organization is oblivious to
'should' or 'ought'.
No doubt. Clearly you failed to appreciate the Microsoft's core
corporate culture: MAKING MONEY COMES FIRST, making decent software is
at best secondary.
Nope. There are lots of strong coders. You're among the few who seem
not to combine (claimed) programming ability with discretion or the
sense to know what's important to your bosses.
Inside Microsoft? Presumably the people who work for Microsoft want to
do so, and most of them realize when to shut up in order to continue to
do so.
Outside Microsoft? Do you read the newsgroups? There are frequent
complaints.
When? I've said I use perl. I've said Python is closely tied into
Gnumeric and OpenOffice Calc. I could add that it's possible to use
both (and REXX and Ruby and any other scripting language capable of
Automation) to automate Excel as a background process (though I'll
admit Excel as a background is a huge waste of resources). And I'll
repeat that there's a perl module that allows perl to create and modify
..XLS files without using Excel.
I've responded to unsubstantiated assertions you've made about how
wide-spread you believe VB, VBA and VBS code is. There may be a fair
amount of in-house VB apps in most medium to large corporations, and
there's definitely a lot of VBA code many extant Office document files.
There doesn't seem to be as much VBS code in use. I can't find any
definitive surveys on the most widely used scripting languages, but
from what digging I have done, it seems clear that a very small
fraction at most of web servers running Apache (so about 70% of all web
servers) would be running VBS code.
All of that's beside the point. Nearly all the people who post in
newsgroups for one or the other of the Office applications are looking
for solutions using those applications. VBA is often but not always
relevant, and it's certainly the best scripting language to use with
Office apps because it's built into the Office framework. VB proper and
VBS are mostly irrelevant for Office apps, though .DLLs could be
written in VB and VBS scripts could be added to Office app-generated
HTML files, but neither of those relatively esoteric tasks are relevant
to the vast majority of postings in Office app newsgroups.
Back to the original topic of this thread. VBA's future lifetime is
independent of that of VB6. Microsoft isn't foolish enough to kill off
VBA any time soon, and it'd offer its successor scripting language
(C#A?) as a parallel option for a few version cycles before it'd drop
VBA. So it's hard to see why people writing strictly VBA code for
Office apps should be any more concerned about the fate of VB6 than
that of COBOL. Similarly for VBS - it's use and future lifecycle are
independent of that of VB6.
So it comes down to whether Microsoft revives a VB6-like programming
language. As I've already written, to in-house VB coders who've already
switched to VB.Net (and they've done so in my company) reviving VB6 is
either an irrelevance or an annoyance. It's no longer an alternative.
Microsoft won't win points with or receive revenues from such customers
by reviving VB6. I have no idea what share of former VB6 programmers
have moved on to VB.Net, but if it's more than half, then A Dios VB6,
it was nice knowin' ya.
I'll also repeat that this is what could happen when one relies on
proprietary software. VB6 programmers who believed that Microsoft would
always be there to support them have no one but themselves to blame for
their own naivete. I won't stand in the way of their petition drive to
keep VB6 alive, but I believe it'll be as efficacious and writing Santa
Claus.
Semantics. No question that they *SHOULD* fix bugs. Lord knows there
are lots of very longstanding bugs in Excel. Problem may be that these
bugs have been around so long that the Excel programmers may consider
them 'features' at this point.
But the only thing Microsoft *NEEDS* to do is earn money. If they can
rake in revenues without fixing bugs, why would they bother? One thing
that must be recognized about Microsoft: no matter how nice & upright
some Microsoft employees may be, Microsoft's corporate culture is, to
be charitable, amoral. Microsoft as an organization is oblivious to
'should' or 'ought'.
No doubt. Clearly you failed to appreciate the Microsoft's core
corporate culture: MAKING MONEY COMES FIRST, making decent software is
at best secondary.
Nope. There are lots of strong coders. You're among the few who seem
not to combine (claimed) programming ability with discretion or the
sense to know what's important to your bosses.
Inside Microsoft? Presumably the people who work for Microsoft want to
do so, and most of them realize when to shut up in order to continue to
do so.
Outside Microsoft? Do you read the newsgroups? There are frequent
complaints.
....
When? I've said I use perl. I've said Python is closely tied into
Gnumeric and OpenOffice Calc. I could add that it's possible to use
both (and REXX and Ruby and any other scripting language capable of
Automation) to automate Excel as a background process (though I'll
admit Excel as a background is a huge waste of resources). And I'll
repeat that there's a perl module that allows perl to create and modify
..XLS files without using Excel.
I've responded to unsubstantiated assertions you've made about how
wide-spread you believe VB, VBA and VBS code is. There may be a fair
amount of in-house VB apps in most medium to large corporations, and
there's definitely a lot of VBA code many extant Office document files.
There doesn't seem to be as much VBS code in use. I can't find any
definitive surveys on the most widely used scripting languages, but
from what digging I have done, it seems clear that a very small
fraction at most of web servers running Apache (so about 70% of all web
servers) would be running VBS code.
All of that's beside the point. Nearly all the people who post in
newsgroups for one or the other of the Office applications are looking
for solutions using those applications. VBA is often but not always
relevant, and it's certainly the best scripting language to use with
Office apps because it's built into the Office framework. VB proper and
VBS are mostly irrelevant for Office apps, though .DLLs could be
written in VB and VBS scripts could be added to Office app-generated
HTML files, but neither of those relatively esoteric tasks are relevant
to the vast majority of postings in Office app newsgroups.
Back to the original topic of this thread. VBA's future lifetime is
independent of that of VB6. Microsoft isn't foolish enough to kill off
VBA any time soon, and it'd offer its successor scripting language
(C#A?) as a parallel option for a few version cycles before it'd drop
VBA. So it's hard to see why people writing strictly VBA code for
Office apps should be any more concerned about the fate of VB6 than
that of COBOL. Similarly for VBS - it's use and future lifecycle are
independent of that of VB6.
So it comes down to whether Microsoft revives a VB6-like programming
language. As I've already written, to in-house VB coders who've already
switched to VB.Net (and they've done so in my company) reviving VB6 is
either an irrelevance or an annoyance. It's no longer an alternative.
Microsoft won't win points with or receive revenues from such customers
by reviving VB6. I have no idea what share of former VB6 programmers
have moved on to VB.Net, but if it's more than half, then A Dios VB6,
it was nice knowin' ya.
I'll also repeat that this is what could happen when one relies on
proprietary software. VB6 programmers who believed that Microsoft would
always be there to support them have no one but themselves to blame for
their own naivete. I won't stand in the way of their petition drive to
keep VB6 alive, but I believe it'll be as efficacious and writing Santa
Claus.
J
Jay Petrulis
'--- snip code ---
OK, you found a link to a matrix inverse function. And the
determinant?
How about the specifics? In the Excel newsgroups (possibly in the
Access newsgroups as well), complete step-by-step solutions are
sometimes given. Please do so for me, if you will.
Set me up from start to finish, please. Granted, that request is more
than is normally required and/or offered, but all you have really done
is given the general outline (again), only this time you added some
code you found. Not enough specifics.
Pretend I just stepped off the Excel boat and onto the promised land.
Convert me to your database-is-everything crusade. I'm still
impressionable, unlike that Excel goon Harlan.
A
aaron.kempf
i dont think that 1/2 of vb6 programmers have moved to .NOT
i dont believe that you or MS has any comprehension of the market.
i see VBS and VBA as _EXACTLY_ the same as vb.
i mean; they're what.. 10% different syntax?
at most?
it's just riduclous that you really think that newbie spreadsheet dorks
should learn perl, python..
you spreadsheet dorks should just lose the training wheels. that's all
im saying.
and there is nothing wrong with investing in vb6-- whether it be vbs or
vba or vb6.
i mean.. it's the best way to write functions.
functions that you can use in
word, excel, access, dts, asp-- you can use vb6-language for many many
things.
you can't reuse java functions in 10 different environments'
you can't reuse perl or python.. you cant' make a function and perl and
paste it into outlook
you can't do it.
i just think that it's funny that you're such an Excel bigot and yet
you dont know jack shit about using excel.
it's like.. kinda reminds me of how hitler was jewish and he was also
anti-jewish
i mean. what do you smoke for breakfast; anyways?
i dont believe that you or MS has any comprehension of the market.
i see VBS and VBA as _EXACTLY_ the same as vb.
i mean; they're what.. 10% different syntax?
at most?
it's just riduclous that you really think that newbie spreadsheet dorks
should learn perl, python..
you spreadsheet dorks should just lose the training wheels. that's all
im saying.
and there is nothing wrong with investing in vb6-- whether it be vbs or
vba or vb6.
i mean.. it's the best way to write functions.
functions that you can use in
word, excel, access, dts, asp-- you can use vb6-language for many many
things.
you can't reuse java functions in 10 different environments'
you can't reuse perl or python.. you cant' make a function and perl and
paste it into outlook
you can't do it.
i just think that it's funny that you're such an Excel bigot and yet
you dont know jack shit about using excel.
it's like.. kinda reminds me of how hitler was jewish and he was also
anti-jewish
i mean. what do you smoke for breakfast; anyways?
A
aaron.kempf
hahah wow Jay i didnt know there were really living breathing thinking
excel dorks.
i'll make a better example tonight.
-aaron
excel dorks.
i'll make a better example tonight.
-aaron
H
Harlan Grove
(e-mail address removed) wrote...
Apparently you don't know Pascal code when you see it. Nor do you
recognize procedure calls that aren't built into the Pascal language so
would also need to be translated from other Pascal code. Looking
through the site to which you provided the link, how quickly would
anyone find the code for lmax, sone, squot, sprod and sdiff?
Also, you don't know that 'simple Gauss-Jordan' means *simple* in the
numerical sense, meaning trouble if you have nearly singular matrices
or several large/small values that could lead to overflow/underflow.
Next, it's built into Excel. You'd have to translate it. Since you
couldn't even identify the language, why should anyone believe you'd do
even a half-assed job translating the code into some language which you
claim to know?
As for write once, to repeat, it's built into Excel, so no need to
reinvent this wheel even via translation. Also, Excel's MINVERSE will
be able to invert more nearly singular matrices which will choke the
simple Gauss-Jordan code you provided. If you want some solid code, see
http://library.lanl.gov/numerical/bookcpdf/c2-3.pdf
If there were procedures I needed repeatedly, I'd write 'em once in VBA
and put them in an .XLA add-in. Then I could call those udfs from any
spreadsheet I use. Gosh! Code reuse in Excel! Whoda thunk?!
But you're missing the key point I made: the relative merits of
spreadsheets and databases are often found in their respective data
referencing features. Given the table I provided in my last post, which
I'll repeat here
Fld1__Fld2__Fld3__Fld4
-7__ -24___ -59___ 58
-78__ 65___ 56___ 22
59__ -60___ -31___ 47
35__ -21___ 10____ 1
how would you pass this matrix to a database udf? And how would the
code below, which uses the passed parameter m2 as the return value,
need to be modified to become a function returning a matrix result? How
would you convert that matrix result back into a table? Cursors?
Quite a few loose ends in your 'complete solution'.
....
Apparently you don't know Pascal code when you see it. Nor do you
recognize procedure calls that aren't built into the Pascal language so
would also need to be translated from other Pascal code. Looking
through the site to which you provided the link, how quickly would
anyone find the code for lmax, sone, squot, sprod and sdiff?
Also, you don't know that 'simple Gauss-Jordan' means *simple* in the
numerical sense, meaning trouble if you have nearly singular matrices
or several large/small values that could lead to overflow/underflow.
Next, it's built into Excel. You'd have to translate it. Since you
couldn't even identify the language, why should anyone believe you'd do
even a half-assed job translating the code into some language which you
claim to know?
As for write once, to repeat, it's built into Excel, so no need to
reinvent this wheel even via translation. Also, Excel's MINVERSE will
be able to invert more nearly singular matrices which will choke the
simple Gauss-Jordan code you provided. If you want some solid code, see
http://library.lanl.gov/numerical/bookcpdf/c2-3.pdf
If there were procedures I needed repeatedly, I'd write 'em once in VBA
and put them in an .XLA add-in. Then I could call those udfs from any
spreadsheet I use. Gosh! Code reuse in Excel! Whoda thunk?!
But you're missing the key point I made: the relative merits of
spreadsheets and databases are often found in their respective data
referencing features. Given the table I provided in my last post, which
I'll repeat here
Fld1__Fld2__Fld3__Fld4
-7__ -24___ -59___ 58
-78__ 65___ 56___ 22
59__ -60___ -31___ 47
35__ -21___ 10____ 1
how would you pass this matrix to a database udf? And how would the
code below, which uses the passed parameter m2 as the return value,
need to be modified to become a function returning a matrix result? How
would you convert that matrix result back into a table? Cursors?
Quite a few loose ends in your 'complete solution'.
L
Lonnie M.
...and now im making 8x as much money as i was back then.
As an employer, I don't believe that I would pay such a narrow-minded
agitator $8 dollars an hour. It would be refreshing to see you go one
post attempting to come up with something other than a four letter word
to emphasize your point. I doubt that you will change the manner in
which you communicate or the value you place on the right tools for the
right job--whether that be spreadsheets or 'novel' languages. That
is why we will always see you as a belligerent one-trick-pony.
Harlin, thank you for your positive contributions to the community
Regards--Lonnie M.
As an employer, I don't believe that I would pay such a narrow-minded
agitator $8 dollars an hour. It would be refreshing to see you go one
post attempting to come up with something other than a four letter word
to emphasize your point. I doubt that you will change the manner in
which you communicate or the value you place on the right tools for the
right job--whether that be spreadsheets or 'novel' languages. That
is why we will always see you as a belligerent one-trick-pony.
Harlin, thank you for your positive contributions to the community
Regards--Lonnie M.
H
Harlan Grove
(e-mail address removed) wrote...
....
First you'd need to translate that Pascal code for inverting matrices,
no? Then you'd have to figure out the table-to-matrix and
matrix-to-table interface. Then as soon as you come across a nasty
nonsingular matrix, you'll need to go find someone else's matrix
inversion code (which you also won't understand), translate it, ensure
your table to udf interface works, then finally be able to start
performing the calculations I would have finished days before.
So sad.
You have a point about #1, but it's a trade off with flexibility. There
are times people use computers standalone. That requires some
redundancy.
As for #2, if it's the *same* formula, how would accuracy differ? Or do
you mean the possibility that inconsistencies may arise in spreadsheet
formulas? That's a real problem, but there are tools available that
sensible Excel users or developers should use to detect and fix it.
And with respect to #3, you've demonstrated that you don't really
understand anything more complicated than counting and summing. I'll be
generous and stipulate that you do a minimally competent job of writing
code implementing procedures for which you're given detailed specs by
people who do know more than grade school math.
....
First you'd need to translate that Pascal code for inverting matrices,
no? Then you'd have to figure out the table-to-matrix and
matrix-to-table interface. Then as soon as you come across a nasty
nonsingular matrix, you'll need to go find someone else's matrix
inversion code (which you also won't understand), translate it, ensure
your table to udf interface works, then finally be able to start
performing the calculations I would have finished days before.
So sad.
....
You have a point about #1, but it's a trade off with flexibility. There
are times people use computers standalone. That requires some
redundancy.
As for #2, if it's the *same* formula, how would accuracy differ? Or do
you mean the possibility that inconsistencies may arise in spreadsheet
formulas? That's a real problem, but there are tools available that
sensible Excel users or developers should use to detect and fix it.
And with respect to #3, you've demonstrated that you don't really
understand anything more complicated than counting and summing. I'll be
generous and stipulate that you do a minimally competent job of writing
code implementing procedures for which you're given detailed specs by
people who do know more than grade school math.
H
Harlan Grove
(e-mail address removed) wrote...
....
I started off only discussing the possible future existence of VB6.
You're the one who conflated that with VBA and VBS. Why don't you try
to stick with the original topic?
Gosh! All statistical analysis boils down to calculating standard
deviations! Whodda thunk? Apparently not all those Excel or SAS or S
Plus dweebs writing time series texts or wasting college and university
resources holding professorial chairs in statistics, quantitative
psychology or econometrics.
To get an idea how shallow your response is, take a peek at
http://herkules.oulu.fi/isbn9514256042/html/x816.html
And for most business PC users, it won't matter that MSDE is free if
they don't already have it on their PCs and they're not allowed to
install nonapproved software.
Your responses remind me of the old joke about three
philosophers/mathematicians/etc in a whole figuring how they'd get out
of it. The first says, "Assume we have a ladder . . ."
....
....
I started off only discussing the possible future existence of VB6.
You're the one who conflated that with VBA and VBS. Why don't you try
to stick with the original topic?
Gosh! All statistical analysis boils down to calculating standard
deviations! Whodda thunk? Apparently not all those Excel or SAS or S
Plus dweebs writing time series texts or wasting college and university
resources holding professorial chairs in statistics, quantitative
psychology or econometrics.
To get an idea how shallow your response is, take a peek at
http://herkules.oulu.fi/isbn9514256042/html/x816.html
And for most business PC users, it won't matter that MSDE is free if
they don't already have it on their PCs and they're not allowed to
install nonapproved software.
Your responses remind me of the old joke about three
philosophers/mathematicians/etc in a whole figuring how they'd get out
of it. The first says, "Assume we have a ladder . . ."
A
aaron.kempf
hey man
it's an easy conversion to vb is all i know
do you really allow people to tell you what kinda perms you have on
your machine?
again; i've told you this before.. maybe these people walk up to you
and see a computer dork so they dont give you perms.
i walk into contracts; and people give me sa/local admin without even
thinking twice. (i think that it has to do with my billrate lol)
it's an easy conversion to vb is all i know
do you really allow people to tell you what kinda perms you have on
your machine?
again; i've told you this before.. maybe these people walk up to you
and see a computer dork so they dont give you perms.
i walk into contracts; and people give me sa/local admin without even
thinking twice. (i think that it has to do with my billrate lol)
A
aaron.kempf
and just for the record; yes.. all math does realistically boil down to
PLUS, MINUS, MULTIPLY AND DIVIDE.
PLUS, MINUS, MULTIPLY AND DIVIDE.
A
aaron.kempf
yeah.. but what are you gonig to do when you hit a measly 64k rows?
i mean.. i poop in 1m row row incremenets
what are you going to do when you need to SHARE this with another user?
email it to them?
what are you going to do when you hit the 2gb limit IN EXCEL? (i've
breached it; have you?)
i mean.. i poop in 1m row row incremenets
what are you going to do when you need to SHARE this with another user?
email it to them?
what are you going to do when you hit the 2gb limit IN EXCEL? (i've
breached it; have you?)
H
Harlan Grove
So you calculate fractional powers and logarithms with power series
approximations?