Skip to content

Instantly share code, notes, and snippets.

@jkopelioff

jkopelioff/Water Sample

Created January 13, 2016 05:47
Show Gist options

  • Select an option

    Learn more about clone URLs
  • Save jkopelioff/84a33e1fa26a496b907b to your computer and use it in GitHub Desktop.

Select an option

Learn more about clone URLs
Save jkopelioff/84a33e1fa26a496b907b to your computer and use it in GitHub Desktop.
Download ZIP
Raw
Water Sample
require 'mysql'
require 'active_record'
ActiveRecord::Base.logger = Logger.new('debug.log')
configuration = YAML::load(IO.read('config/database.yml'))
ActiveRecord::Base.establish_connection(configuration['development'])
class FactorWeight < ActiveRecord::Base
end
#
# Assignment:
#
# Implement the methods as specified in the following class, plus any you need to use to make your life easier.
# Make explicit any dependencies on external libraries so that I can run your code locally.
#
# Los Angeles recently sampled water quality at various sites, and recorded the
# presence of contaminants. Here's an excerpt of the table:
# (from: http://file.lacounty.gov/bc/q3_2010/cms1_148456.pdf)
# (All chemical values are in mg/L)
# | id | site | chloroform | bromoform | bromodichloromethane | dibromichloromethane |
# | 1 | LA Aqueduct Filtration Plant Effluent | .00104 | .00000 | .00149 | .00275 |
# | 2 | North Hollywood Pump Station (well blend) | .00291 | .00487 | .00547 | .0109 |
# | 3 | Jensen Plant Effluent | .00065 | .00856 | .0013 | .00428 |
# | 4 | Weymouth Plant Effluent | .00971 | .00317 | .00931 | .0116 |
#
# These four chemical compounds are collectively regulated by the EPA as Trihalomethanes,
# and their collective concentration cannot exceed .080 mg/L
#
# (from http://water.epa.gov/drink/contaminants/index.cfm#List )
class WaterSample < ActiveRecord::Base
# This class intends to ease the managing of the collected sample data,
# and assist in computing factors of the data.
#
# The schema it must interact with and some sample data should be delivered
# with your assignment as a MySQL dump
def self.find(sample_id)
# spec
# sample2 = WaterSample.find(2)
# sample2.site.should == "North Hollywood Pump Station (well blend)")
# sample2.chloroform.should == 0.00291
# sample2.bromoform.should == 0.00487
# sample2.bromodichloromethane.should == 0.00547
# sample2.dibromichloromethane.should == 0.0109
# rs = @@con.query "SELECT * FROM water_samples WHERE id = #{sample_id} LIMIT 1"
# new(rs.fetch_hash)
super
end
# Some Trihalomethanes are nastier than others, bromodichloromethane and
# bromoform are particularly bad. That is, water that has .060 mg/L of
# Bromoform is generally more dangerous than water that has .060 mg/L of
# Chloroform, though both are considered "safe enough".
#
# We could build a better metric by adjusting the contribution of each
# component to the Trihalomethane limit based on it's relative "danger".
# Furthermore, consider we want to try several different combinations of
# weights (factors).
#
# Sample table of Factor weights:
# | id | chloroform_weight | bromoform_weight | bromodichloromethane_weight | dibromichloromethane_wieight |
# | 1 | 0.8 | 1.2 | 1.5 | 0.7 |
# | 2 | 1.0 | 1.0 | 1.0 | 1.0 |
# | 3 | 0.9 | 1.1 | 1.3 | 0.6 |
# | 4 | 0.0 | 1.0 | 1.0 | 1.7 |
#
# In statistics, a factor is a single value representing a combination of
# several component values. We may gather several different variables, which
# semantically indicate a similar idea, and, to make analysis simpler, we can
# combine these several values into a single "factor" and disregard the
# constituents
#
# The weights that we should use in our factor could be a complex question.
# Ultimately it depends on what we're modeling.
#
# For example, let's say the city has the option of installing one of several
# different filtration units to remove a specific Triahlomethane (but the city
# can't afford all of the filters). We can use differently weighted factors to
# simulate each of these and do a cost / benefit analysis informing the city's
# decision on which filtration unit to purchase.
#
# Let's say someone from the city has already computed various factors they
# want to analyze, and put them in a factor_weights table.
# In our case, we'll use a linear combination of the nth factor weights to compute the
# samples nth factor.
#
#
#
# Return the value of the computed factor with id of factor_weights_id
def factor(factor_weights_id)
# spec:
# sample2 = WaterSample.find(2)
# sample2.factor(6) #computes the 6th factor of sample #2
# => .0213
# Note that the factor for this example is from data not in the sample data
# above, that's because I want you to be sure you understand how to compute
# this value conceptually.
factor_weight = FactorWeight.find(factor_weights_id)
return attributes.except("id", "site").sum { |attribute, value| value * factor_weight["#{attribute}_weight"] }.round(4)
end
# convert the object to a hash
# if include_factors is true, include all computed factors in the hash
def to_hash(include_factors = false)
# spec:
# sample2.to_hash
# => {:id =>2, :site => "North Hollywood Pump Station (well blend)", :chloroform => .00291, :bromoform => .00487, :bromodichloromethane => .00547 , :dibromichlormethane => .0109}
# sample2.to_hash(true)
# #let's say only 3 factors exist in our factors table, with ids of 5, 6, and 9
# => {:id =>2, :site => "North Hollywood Pump Station (well blend)", :chloroform => .00291, :bromoform => .00487, :bro modichloromethane => .00547 , :dibromichlormethane => .0109, :factor_5 => .0213, :factor_6 => .0432, :factor_9 => 0.0321}
result = attributes
if include_factors
factor_weight_ids = FactorWeight.pluck(:id)
factor_weight_ids.inject(result) do |hash, id|
hash["factor_#{id}"] = factor(id)
hash
end
end
return result.symbolize_keys
end
end
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment