Linear regression trendline calculator

Require the gem as per below. We'll require a few more for the notebook



In [1]:

    
require 'pry'
require 'awesome_print'
require 'nyaplot'

require 'linear_regression_trend'









    














    Out[1]:





true

Usage

Have an array of samples handy, create a trend object like so:

trender = LinearRegressionTrend::Calculator.new(my_array_of_samples)

trended_samples = trender.trend
trend_slope     = trender.slope
trend_intercept = trender.intercept

See the below example for more examples, and a visualisation.



In [2]:

    
y = [1,2,3,6,10,11,12,12,12,12,15,18,18,16,16,15,14,11,9,9,7,4,2,1,1]
x = (1..y.size).to_a.map { |n| n.to_s }

t = LinearRegressionTrend::Calculator.new(y)

IRuby.display IRuby.table([['Slope', 'Intercept'], [t.slope, t.intercept]])









    




Slope Intercept
-0.06384615384615384 10.31

Given a known set of samples



In [3]:

    
chart = Nyaplot::Plot.new 
chart.add(:bar, x, y)
chart.show

The trendline for that sample set should be a descending trend



In [4]:

    
chart = Nyaplot::Plot.new 
chart.add(:bar, x, t.trend)
chart.show

Non-negative descending trends



In [13]:

    
y = [1,2,3,0,10,18,18,18,1,1,1,1,0.5,0.5,0.5,0.5,05,0.25,0.25,0.1,0.1,0,0,0,0]
x = (1..y.size).to_a.map { |n| n.to_s }

t = LinearRegressionTrend::Calculator.new(y, non_negative: true)

IRuby.display IRuby.table([['Slope', 'Intercept', 'Conclusion'], [t.slope, t.intercept, t.trend.last]])









    




Slope Intercept Conclusion
-0.3467307692307691 7.775499999999997 0



In [12]:

    
chart = Nyaplot::Plot.new 
chart.add(:bar, x, t.trend)
chart.show

It's last few values should be 0, instead of negative



In [10]:

    
IRuby.display IRuby.table([(1..y.size).to_a.reverse, t.trend.reverse])









    




25 24 23 22 21 20 19 … 8 7 6 5 4 3 2
0 0 0 0.14742307692307666 0.4941538461538455 0.8408846153846152 1.187615384615384 5.001653846153845 5.348384615384614 5.695115384615383 6.041846153846151 6.388576923076921 6.73530769230769 7.082038461538459

Slope	Intercept	Conclusion
-0.3467307692307691	7.775499999999997	0

25	24	23	22	21	20	19	…	8	7	6	5	4	3	2
0	0	0	0.14742307692307666	0.4941538461538455	0.8408846153846152	1.187615384615384	…	5.001653846153845	5.348384615384614	5.695115384615383	6.041846153846151	6.388576923076921	6.73530769230769	7.082038461538459