Wouldn't it just be easier to slurp the data into Access, and export an Excel spreadsheet? Methinks, yes, having had to do way too many Excel-based hacks that just work better in Access (or Sql Server, Oracle, etc).
dim fn as string dim path as string path="C:\*.txt"
fn = dir(path) 'gets the first file name that matches the pattern
You are talking about the physical limitations of the table. It says nothing about the actual set up of the data. A pivot table is denormalized data.
It is just like saying arrays in any programming language are just a lazy syntactical trick, because they usually are mapped to a linear memory space anyways, and look what happens when you run this row-centric code vs column-centric code with this compiler.
The different field values in the table are the dimensions. So it's a hack to fit it into a two-dimensional storage space. BFD. It's no more of a hack that C provides to do a 3-dimesional array in a linear memory model.
Look a little bit deeper in your manuals about the "WITH CUBE" group by option (SS7, Ora8+). This is more what a pivot table provides. Internally it has the calculations for all combinations of the "dimension" columns stored in it, so that instead of running through the whole table to get a different analysis, it just pulls the relevant calculated fields for the desired sets of dimension fields.
As far as adding dimensions ad infinitum, it's easy enough to do it in SQL as well. Just add more dimension fields (and dimension tables) to the database.
Besides, for the scenario you talk about, SS7/2K, Oracle 8i/9i/10g, etc., let you "partition" tables on things like this. You make lots of separate tables, build a UNION query that ties them all together, and add constraints to the field they're separated by so that the query optimizer can look and quickly figure out that it only needs to hit a smaller partitioned table to get the data, rather than looking in the much larger whole tableset.
Oracle is nicer, because you can set up the partitioning to be essentially invisible, and much less specific, along the lines of (partition a = this month's data, part. B = the prev. 3 months before that, Part. c = the rest of the year, and Part. D -... = years before that). Sql Server is not very abstract like that for partitioning schemes.
But this partitioning doesn't make the data "multi-dimensional". Sticking the year field into the table makes it multi-dimensional.
OK, Mr. Lameass reader, do you also need a tutorial on how to start up IE, type in "google.com" in the address bar, and do a Google search for "excel pivot table"?
SQL tables can be considered a one-dimensional view of N-dimensional data. You can, through the right query, produce a view based on any or or two of those N dimensions. Only the two-dimensional ones report a pivot table/matrix report....yes, and no. It depends on the source.
select country, industry_grp, year, sum(population) pop_sum, sum(gdp) gdp_sum into census_cube from census_table group by country, industry_grp, year WITH CUBE...is a multi-dimensional table... The original table was multidimensional too (country, year dimensions).
What you are saying is that "a multi-dimensional array isn't really multi-dimensional, because it's mapped to a linear memory space". It's all true at the physical level, but in practice, it's a multi-dimensional structure.
Like the pivot table, the goal is to store (all of) the summary results into a storage structure (server table, pivot table memory), to hasten analysis so that query time over the parent dataset is reduced to a single action, and retrieving the summary results goes from walking through a 100K (or 1M) transactional table to querying a much smaller table that already has your results.
This is sort of the point anyways of a data warehouse, products like Hyperion, Business Objects, etc.
An Access cross-tab query can only have one "column heading" field.
But Cross-tab queries are workable to help denormalize textual data.
Pivot tables are only for statistical analysis on numeric data.
Pivot tables are also only really useful when pulling data from MS-Query or OLAP cubes. The 64K-row limit in an Excel worksheet precludes using these as the basis for heavy-duty pivot table work.
What is the underlying data look like? Well, if you have a big summary query, and your SQL implementation supports the WITH CUBE statements (i.e., select x, y, z, sum(z) from the_table group by x, y, z WITH CUBE), then the output from running one of these is stored in memory. Rather than cycle through the dataset to resummarize the data, it has all the summaries already in it, and you really end up just pulling the various summary values.
Sure, Pivot tables will do the recalc occaisionally (like if you change your summary function), but it's in memory, and it's generally fast, much faster than rerunning the query.
Because it's in-memory, if you're hitting a million-row transactional table, you're probably gonna need a lot of RAM to work with it...
Re:Better than a Volcano
on
Hacking Vodka
·
· Score: 1
it's not physically possible to distill all of the water out of EtOH (has to do with phase change overlaps, vapor pressure/temperatures, etc. between water and EtOH).
You can always buy a bottle of Everclear (194 proof, 97% EtOH, about as close to economical pure distilled EtOH that you can get).
The benzene, I would think, binds with either the EtOH or H2O, to mess up the relative densities between the two, so that the EtOH can be more easily separated. That of course would leave some residual benzene in the EtOH... which, like was said, is probably not good to ingest...
For the earth space elevator, I'm thinking that short of a nuculer bomb being detonated at its base, it's going to be pretty damn near impervious, even to a loaded 747-400 crashing into it at 600 mph.
Think of the main cables for a suspension bridge. Those definitely are not the weak points of a suspension bridge...
No, carbon nanotubes are being proposed because of their high tensile strength and, I assume, very low elasticity. The mass on the end will provide the "pulling" force.
We've already done the "tether small object from big object", with a couple of tethered satellite experiments from the space shuttle.
But you would think that to "commercialize" the moon, you would first need an economically viable way to get stuff up and down to the earth, something far more economical than blasting it up on an Arianne 5, SS-18 or Delta IV, and letting it drop into the ocean from LEO.
This analysis is WAY beyond putting the cart in front of the horse.
Doesn't France have the largest islamic population in Europe, at something like 15%? Seems like Chirac was just doing some political ass-saving as well as keeping various contracts (oil, military, construction, etc) positive for France.
When I was at the Univ. of Washington, there was no love lost between the Korean students and Japanese students... It just wasn't brought up much at all as a topic for conversation by anybody.
Japanese can insist they can tell gaijin Japanese as well (full-blood Japanese who have lived outside of Japan for too long).
Oh well, probably no different than living in Greece and not being Greek Orthodox...
Don't forget the physical damage that all those high energy neutrons cause over time, degrading the strength of the materials in the containment vessel.
This is one of the prime factors limiting the lifetime of a fission nuclear reactor pressure vessel. After about 20 years absorbing a huge neutron flux, even 10" thick steel, I suppose, gets weak enough that 4000psi steam rupturing it becomes a concern.
So on the day I agree to the oil futures contract I also purchase a dollar/euro forward contract such that I lock in an exchange rate where I pay x euros and receive y dollars, where y is the amount of the oil futures contract.
PRESTO! I have just traded oil in euros.
Now you could make the argument that increased trade on the dollar provides the dollar with a liquidity premium and that is true. But it isn't as big as one might think.
Even better, you let your financial exchange arbitrage people manage the timing of the contracts, so that not only do you profit from the petro future, you benefit greatly from currency exchange fluctuations.
Ideally, you the US dollars with Euros when it is advantageous (high Euro::USD ratio), and you convert back to Euros when it's advantageous (lower Euro::USD ratio).
Sorry, most ruminant burps are pretty methane-laced as well... most of the breakdown happens in the rumen. If you have ever had a sheep or cow burp in your face, you would understand.
Who cares if the tech is open-sourced? You still have to throw a few $billion at ABB, Westinghouse, GE Power Systems, Halliburton, etc. to actually physically build the plant, comply with environmental regulations, build up the electrical system, etc.
It's like worrying about a few $thousand over OS prices when buying a big ol' SunFire 64-way server.
How exactly does solar help provide power nearly half of the time?
It might be good to help provide power during daylight hours, but not to provide electricity to the US northeast in the winter, where it's usually cold during the day and colder during the night.
Besides, if the US started giving solar power systems to foreign countries, the EU will start making noise in the WTO about yet another subsidy to an American industry by the US government that makes it compete less well with European companies that have to sell said tech.
It's the 5-10 asshats following you through the intersection AFTER the light has turned red that partially gum it up.
Then, say the first "legit" car has to stop before they make it through the interesction, because the traffic is backed up all the way to the intersection. Now, you have ALL lane directions gummed up until all those other cars have made it through. Aero Drive just west of I-15 in San Diego is one, especially this time of year. Everyone's gotta get to Wal-Mart to buy that Pickle-Me-Elmo before someone else does!
One study I read showed that rural taxpayers used $0.70 per tax dollar they contributed (state? county?) compared to urban tax payer that used $1.40 per tax dollar contributed.
If a state rule mandating tax money stay in county it's collected in were to be mandated, I suspect the state would move to diminish the number of counties. If you live in Texas or California, it would be interesting, to say the least. There are two counties between the Pacific Ocean and Yuma, AZ. Riverside County is also pretty dang big as well.
IN areas like San Diego, Puget Sound, Chicagoland, etc., the existing shitty traffic doesn't seem to be doing anything at all at limiting urban sprawl, so what's your point, that Portland now gets to be like the rest of the country's major cities, thanks to Prop 37?
My wife worked with people who commuted to San Diego Naval Hospital from Temecula, about 60 miles away. "you get a bigger house for your money in Temecula!"
Sorry, road maintenance still requires massive tax infusions as well. Where does most gas tax $$$ go to? Highway maintenance. So, less gas use, less use of roads, less use of tax $$$ for roads. How much does Ill. Dept. of Transportation spend on I-94 maintenance per passenger-mile compared to Metra, for example?
Sounds like the busses I used to catch: Metro 307 to Bothell Park & Ride, xfer to CT 120 to Bothell/Mill Creek.
Sucked big time to see the damn 120 leaving the P&R as the 307 was just pulling up to it in the morning, necessitating a 45 minute wait for the next one, so half the time I walked the 4 miles to where I needed to go.
even though there is no real need to make your own sweaters anymore.
There is if you want one made out of yak wool, alpaca, dog wool, belly button lint, etc.
Why mention alpaca? It takes about 1 lb to make enough alpaca yarn to make a sweater. At the store, said sweater will cost at least $100. It came from one of the cooperative mills in Peru or Bolivia, which cost the coop about $5.00 to make.
So at the expense of a few days' knitting, and maybe $20 in yarn and needles you can make your own fuzzy alpaca sweater.
Besides, why do you use Linux, when Microsoft will take care of you just fine with Windows and all its vaunted support and superior innovation?
Slashdot Mirror
I feel sorry for you.
Wouldn't it just be easier to slurp the data into Access, and export an Excel spreadsheet? Methinks, yes, having had to do way too many Excel-based hacks that just work better in Access (or Sql Server, Oracle, etc).
dim fn as string
dim path as string
path="C:\*.txt"
fn = dir(path) 'gets the first file name that matches the pattern
while fn ""
'do something with fn
dir() 'gets the next file name match
wend
=sumif(range-for-condition, condition-value, range-to-sum)
..etc.
ex: =sumif("$a:$a","between 1990 and 2000", "$d:$d")
=countif()
Array formulas are slightly different.
You are talking about the physical limitations of the table. It says nothing about the actual set up of the data. A pivot table is denormalized data.
... = years before that). Sql Server is not very abstract like that for partitioning schemes.
It is just like saying arrays in any programming language are just a lazy syntactical trick, because they usually are mapped to a linear memory space anyways, and look what happens when you run this row-centric code vs column-centric code with this compiler.
The different field values in the table are the dimensions. So it's a hack to fit it into a two-dimensional storage space. BFD. It's no more of a hack that C provides to do a 3-dimesional array in a linear memory model.
Look a little bit deeper in your manuals about the "WITH CUBE" group by option (SS7, Ora8+). This is more what a pivot table provides. Internally it has the calculations for all combinations of the "dimension" columns stored in it, so that instead of running through the whole table to get a different analysis, it just pulls the relevant calculated fields for the desired sets of dimension fields.
As far as adding dimensions ad infinitum, it's easy enough to do it in SQL as well. Just add more dimension fields (and dimension tables) to the database.
Besides, for the scenario you talk about, SS7/2K, Oracle 8i/9i/10g, etc., let you "partition" tables on things like this. You make lots of separate tables, build a UNION query that ties them all together, and add constraints to the field they're separated by so that the query optimizer can look and quickly figure out that it only needs to hit a smaller partitioned table to get the data, rather than looking in the much larger whole tableset.
Oracle is nicer, because you can set up the partitioning to be essentially invisible, and much less specific, along the lines of (partition a = this month's data, part. B = the prev. 3 months before that, Part. c = the rest of the year, and Part. D -
But this partitioning doesn't make the data "multi-dimensional". Sticking the year field into the table makes it multi-dimensional.
OK, Mr. Lameass reader, do you also need a tutorial on how to start up IE, type in "google.com" in the address bar, and do a Google search for "excel pivot table"?
SQL tables can be considered a one-dimensional view of N-dimensional data. You can, through the right query, produce a view based on any or or two of those N dimensions. Only the two-dimensional ones report a pivot table/matrix report. ...yes, and no. It depends on the source.
...is a multi-dimensional table... The original table was multidimensional too (country, year dimensions).
select country, industry_grp, year, sum(population) pop_sum, sum(gdp) gdp_sum
into census_cube
from census_table
group by country, industry_grp, year
WITH CUBE
What you are saying is that "a multi-dimensional array isn't really multi-dimensional, because it's mapped to a linear memory space". It's all true at the physical level, but in practice, it's a multi-dimensional structure.
Like the pivot table, the goal is to store (all of) the summary results into a storage structure (server table, pivot table memory), to hasten analysis so that query time over the parent dataset is reduced to a single action, and retrieving the summary results goes from walking through a 100K (or 1M) transactional table to querying a much smaller table that already has your results.
This is sort of the point anyways of a data warehouse, products like Hyperion, Business Objects, etc.
You mean, an Access Cross-tab query? Yes and no.
An Access cross-tab query can only have one "column heading" field.
But Cross-tab queries are workable to help denormalize textual data.
Pivot tables are only for statistical analysis on numeric data.
Pivot tables are also only really useful when pulling data from MS-Query or OLAP cubes. The 64K-row limit in an Excel worksheet precludes using these as the basis for heavy-duty pivot table work.
What is the underlying data look like? Well, if you have a big summary query, and your SQL implementation supports the WITH CUBE statements (i.e., select x, y, z, sum(z) from the_table group by x, y, z WITH CUBE), then the output from running one of these is stored in memory. Rather than cycle through the dataset to resummarize the data, it has all the summaries already in it, and you really end up just pulling the various summary values.
Sure, Pivot tables will do the recalc occaisionally (like if you change your summary function), but it's in memory, and it's generally fast, much faster than rerunning the query.
Because it's in-memory, if you're hitting a million-row transactional table, you're probably gonna need a lot of RAM to work with it...
it's not physically possible to distill all of the water out of EtOH (has to do with phase change overlaps, vapor pressure/temperatures, etc. between water and EtOH).
You can always buy a bottle of Everclear (194 proof, 97% EtOH, about as close to economical pure distilled EtOH that you can get).
The benzene, I would think, binds with either the EtOH or H2O, to mess up the relative densities between the two, so that the EtOH can be more easily separated. That of course would leave some residual benzene in the EtOH... which, like was said, is probably not good to ingest...
For the earth space elevator, I'm thinking that short of a nuculer bomb being detonated at its base, it's going to be pretty damn near impervious, even to a loaded 747-400 crashing into it at 600 mph.
Think of the main cables for a suspension bridge. Those definitely are not the weak points of a suspension bridge...
No, carbon nanotubes are being proposed because of their high tensile strength and, I assume, very low elasticity. The mass on the end will provide the "pulling" force.
We've already done the "tether small object from big object", with a couple of tethered satellite experiments from the space shuttle.
But you would think that to "commercialize" the moon, you would first need an economically viable way to get stuff up and down to the earth, something far more economical than blasting it up on an Arianne 5, SS-18 or Delta IV, and letting it drop into the ocean from LEO.
This analysis is WAY beyond putting the cart in front of the horse.
Algeria.
I'll leave it at that. Au revoir.
Doesn't France have the largest islamic population in Europe, at something like 15%? Seems like Chirac was just doing some political ass-saving as well as keeping various contracts (oil, military, construction, etc) positive for France.
When I was at the Univ. of Washington, there was no love lost between the Korean students and Japanese students... It just wasn't brought up much at all as a topic for conversation by anybody.
Japanese can insist they can tell gaijin Japanese as well (full-blood Japanese who have lived outside of Japan for too long).
Oh well, probably no different than living in Greece and not being Greek Orthodox...
Yes, the Russian CV flies navalized version of MiG-29...
Don't forget the physical damage that all those high energy neutrons cause over time, degrading the strength of the materials in the containment vessel.
This is one of the prime factors limiting the lifetime of a fission nuclear reactor pressure vessel. After about 20 years absorbing a huge neutron flux, even 10" thick steel, I suppose, gets weak enough that 4000psi steam rupturing it becomes a concern.
Why doesn't the EU simply name Switzerland as the lead nation, with France as the prime contractor?
If Airbus can "join" with Sikorsky to try and sell Eurocopters as an "american" helicopter, why not?
So on the day I agree to the oil futures contract I also purchase a dollar/euro forward contract such that I lock in an exchange rate where I pay x euros and receive y dollars, where y is the amount of the oil futures contract.
PRESTO! I have just traded oil in euros.
Now you could make the argument that increased trade on the dollar provides the dollar with a liquidity premium and that is true. But it isn't as big as one might think.
Even better, you let your financial exchange arbitrage people manage the timing of the contracts, so that not only do you profit from the petro future, you benefit greatly from currency exchange fluctuations.
Ideally, you the US dollars with Euros when it is advantageous (high Euro::USD ratio), and you convert back to Euros when it's advantageous (lower Euro::USD ratio).
carbon dioxide comes from burps
Sorry, most ruminant burps are pretty methane-laced as well... most of the breakdown happens in the rumen. If you have ever had a sheep or cow burp in your face, you would understand.
Who cares if the tech is open-sourced? You still have to throw a few $billion at ABB, Westinghouse, GE Power Systems, Halliburton, etc. to actually physically build the plant, comply with environmental regulations, build up the electrical system, etc.
It's like worrying about a few $thousand over OS prices when buying a big ol' SunFire 64-way server.
How exactly does solar help provide power nearly half of the time?
It might be good to help provide power during daylight hours, but not to provide electricity to the US northeast in the winter, where it's usually cold during the day and colder during the night.
Besides, if the US started giving solar power systems to foreign countries, the EU will start making noise in the WTO about yet another subsidy to an American industry by the US government that makes it compete less well with European companies that have to sell said tech.
It's the 5-10 asshats following you through the intersection AFTER the light has turned red that partially gum it up.
Then, say the first "legit" car has to stop before they make it through the interesction, because the traffic is backed up all the way to the intersection. Now, you have ALL lane directions gummed up until all those other cars have made it through. Aero Drive just west of I-15 in San Diego is one, especially this time of year. Everyone's gotta get to Wal-Mart to buy that Pickle-Me-Elmo before someone else does!
Go to San Diego or LA to see this in action.
Hmm... rural vs urban tax usage...
One study I read showed that rural taxpayers used $0.70 per tax dollar they contributed (state? county?) compared to urban tax payer that used $1.40 per tax dollar contributed.
If a state rule mandating tax money stay in county it's collected in were to be mandated, I suspect the state would move to diminish the number of counties. If you live in Texas or California, it would be interesting, to say the least. There are two counties between the Pacific Ocean and Yuma, AZ. Riverside County is also pretty dang big as well.
IN areas like San Diego, Puget Sound, Chicagoland, etc., the existing shitty traffic doesn't seem to be doing anything at all at limiting urban sprawl, so what's your point, that Portland now gets to be like the rest of the country's major cities, thanks to Prop 37?
My wife worked with people who commuted to San Diego Naval Hospital from Temecula, about 60 miles away. "you get a bigger house for your money in Temecula!"
Sorry, road maintenance still requires massive tax infusions as well. Where does most gas tax $$$ go to? Highway maintenance. So, less gas use, less use of roads, less use of tax $$$ for roads. How much does Ill. Dept. of Transportation spend on I-94 maintenance per passenger-mile compared to Metra, for example?
Sounds like the busses I used to catch: Metro 307 to Bothell Park & Ride, xfer to CT 120 to Bothell/Mill Creek.
Sucked big time to see the damn 120 leaving the P&R as the 307 was just pulling up to it in the morning, necessitating a 45 minute wait for the next one, so half the time I walked the 4 miles to where I needed to go.
even though there is no real need to make your own sweaters anymore.
There is if you want one made out of yak wool, alpaca, dog wool, belly button lint, etc.
Why mention alpaca? It takes about 1 lb to make enough alpaca yarn to make a sweater. At the store, said sweater will cost at least $100. It came from one of the cooperative mills in Peru or Bolivia, which cost the coop about $5.00 to make.
So at the expense of a few days' knitting, and maybe $20 in yarn and needles you can make your own fuzzy alpaca sweater.
Besides, why do you use Linux, when Microsoft will take care of you just fine with Windows and all its vaunted support and superior innovation?