search

fasterthanlime / cs322

CS-323 project
2 stars 1 forks source link

Implement query R #27

Closed nddrylliog closed 12 years ago

nddrylliog commented 12 years ago

List the best 10 schools for each of the following categories: scorers, rebounders, blockers. Each school’s category ranking is computed as the average of the statistical value for 5 best players that went to that school. Use player’s career average for inputs.

nddrylliog commented 12 years ago

O_O How the fuck would we do that.

That's' 3 different queries!

greut commented 12 years ago

It is. What scares you?

Create a JOIN between players (the denormalized table) and locations, keep the top 5 for each location (hint: ROW_NUMBER() OVER PARTITION), compute the avg value and sort that. Piece of cake… almost.

If you want to denormalize the location_id into the player, I think it's doable and actually make sense. But I'll let you do it so you get a chance to play with all that power once.

nddrylliog commented 12 years ago

what I'm saying is: it doesn't make sense to display it all in the same table/query/view, we won't be able to know by which criterion a school is 'the best'

greut commented 12 years ago

According to the description, one table for each category seems decent. Or a drop down to select which criterion you want to compare them on. Mixing everything up doesn't make sense, maybe a 3D graph using WebGL but it's not in the scope of this project ;-)

nddrylliog commented 12 years ago

See, so that was my original worry. I can work with that.

On Sat, May 26, 2012 at 5:57 PM, Yoan Blanc < reply@reply.github.com

wrote:

According to the description, one table for each category seems decent. Or a drop down to select which criterion you want to compare them on. Mixing everything up doesn't make sense, maybe a 3D graph using WebGL but it's not in the scope of this project ;-)

Reply to this email directly or view it on GitHub: https://github.com/nddrylliog/cs322/issues/27#issuecomment-5947804

Amos Wenger — amos.me Engineering official.fm, founder thebootstrap.ch I'm @nddrylliog https://github.com/nddrylliog on GitHub

greut commented 12 years ago

I'll have to use the same ROW_NUMBER OVER (PARTITION …) trick for the query Q #26. It looks dirty but doable so far, BBQ-time now. Good evening!

nddrylliog commented 12 years ago

Enjoy the barbecue :)

On Sat, May 26, 2012 at 6:22 PM, Yoan Blanc < reply@reply.github.com

wrote:

I'll have to use the same ROW_NUMBER OVER (PARTITION …) trick for the query Q #26. It looks dirty but doable so far, BBQ-time now. Good evening!

Reply to this email directly or view it on GitHub: https://github.com/nddrylliog/cs322/issues/27#issuecomment-5947946

Amos Wenger — amos.me Engineering official.fm, founder thebootstrap.ch I'm @nddrylliog https://github.com/nddrylliog on GitHub

greut commented 12 years ago

Find the top 5 example there: 4627ea9d25e1abd88b6521ba4b49efecdbbfa129

nddrylliog commented 12 years ago

@greut Did you test that? I"m on the train to SwissJeese, but my computer at home on which I usually SSH didn't seem to connect to the network, hence I can't work :(

greut commented 12 years ago

@sebasti4n can you please finalize the view (the r.html.erb file is missing) and put a couple of line in the report as well.

SeZuo commented 12 years ago

all right!

greut commented 12 years ago

I've commented in the commits themselves… imho it's far from being ok.

greut commented 12 years ago

À deux cent lieux d'être dans le mille.