Ignoring nil parameters in array select - ruby

Question

Example
The user has made a search and has specified an age but not a number of children, i.e. wants to find people with any number of children, including zero.
array = []
obj1 = {:name => "Steve", :age => 32, :children => 2}
obj2 = {:name => "Dave", :age => 37, :children => 4}
obj3 = {:name => "Barry", :age => 40, :children => 0}
array << obj1
array << obj2
array << obj3
puts array
## replicating incoming search parameters
params = {}
params[:age] = 35
params[:children] = nil
matching_params = array.select{|person| person[:age] > params[:age] && person[:children] > params[:children]}
If I run this code, I will get an error that it can't compare a number with nil.
Workaround
If I change the children part to && person[:children] > (params[:children] || -1) then this will show all people over 35. For this situation, it does what's required.
Problem
Imagine, however, a world where it's possible to have a negative number of children...I would have to change -1 to minus infinity. Is there some way to just exclude any search criteria which have nil values?

Answer 1

Explicitly filter in nils.
array = [
{:name => "Steve", :age => 32, :children => 2},
{:name => "Dave", :age => 37, :children => 4},
{:name => "Barry", :age => 40, :children => 0}
]
params = {age: 35, children: nil}
matching_params =
array.select do |person|
(params[:age].nil? || person[:age] > params[:age]) &&
(params[:children].nil? || person[:children] > params[:children])
end

Answer 2

Code
def select_by_age_and_children(arr, params)
arr.select do |p|
p[:age] > (params[:age] || -1) &&
p[:children] > (params[:children] || -1)
end
end
Examples
array = [
{:name => "Steve", :age => 32, :children => 2},
{:name => "Dave", :age => 37, :children => 4},
{:name => "Barry", :age => 40, :children => 0}
]
params = { age: 35, children: 1 }
select_by_age_and_children(array, params)
#=> [{:name=>"Dave", :age=>37, :children=>4}]
params = { age: 35, children: nil }
select_by_age_and_children(array, params)
#=> [{:name=>"Dave", :age=>37, :children=>4},
# {:name=>"Barry", :age=>40, :children=>0}]
params = { age: 0, children: 1 }
select_by_age_and_children(array, params)
#=> [{:name=>"Steve", :age=>32, :children=>2},
# {:name=>"Dave", :age=>37, :children=>4}]
params = { age: nil, children: nil }
select_by_age_and_children(array, params)
#=> [{:name=>"Steve", :age=>32, :children=>2},
# {:name=>"Dave", :age=>37, :children=>4},
# {:name=>"Barry", :age=>40, :children=>0}]
Alternative design
Consider writing the method as follows:
def select_by_age_and_children(arr, params)
arr.select { p[:age] >= params[:age] && p[:children] >= params[:children] }
end
and then instead of calling it with, say:
params = { age: 35, children: nil }
call it with:
params = { age: 36, children: 0 }
As well as simplifying the code this avoids the need for the reader to understand what is meant by a nil value in params.