Rails, finder in model (using collect) or in controller (iterating and using push). Which is more efficent?

Question

which way below is more efficient to accomplish this task. I want to loop through a model, check if a list made up of integer vals associated with each model_id is greater than 0. If it is then take the corresponding models into a list of models.

@models = Model.find(:all).collect{|m| m }.reject{ |i| modellist[i.id] < 1 }

or like this

finalModels = []
Model.find_each do |model|
  if modellist[model.id] > 0 #edited
  #if modellist[model.id] != 0
    finalModels.push( model )
  end
end
@models = finalModels

Im leaning towards the second approach, but im not sure. Maybe some insight into how .collect and .reject works to see how efficient it is.

• Edit

My model is called Picture. modellist (or pList) contains data similar to this.

[nil,nil,nil,3,nil,nil,nil,nil,nil,nil,nil,nil,nil,nil,
nil,nil,7,nil,nil,nil,0,nil,nil,nil,0,0,nil,nil,1,3]

I the index number of pList corresponds to the picture id for this. So i need to find pics where pList[picture id] is greater than 0.

• Edit

Used Benoit Garrets answer. What i had to do was make sure that pList was declared by pList = Hash.new and not pList = []. the exact query i used was

@pictures = Picture.find(pList.select {|k, v| v > 0}.keys)

Answer 1

You can just do this in the database directly:

@models = Model.find(:all, :conditions => ["id in (?)", modellist.select{ |i| i && i > 0}])

Or the shorter version:

@models = Model.find(modellist.select{ |i| i && i > 0})

Answer 2

Assuming your modellist and pList already exist and are Hash, you could filter them before and use find with an array:

@models = Model.find(modellist.reject {|k, v| v < 1} .keys)

Same thing with your second example:

@pictures = Picture.find(pList.reject {|k, v| v > 0} .keys)

This way, you won't be looping over your entire database.