Frusterations With Phind

imaynotusephind.md

I don't personally think Phind is ready for offering refinements for code, for instance I offered an example of why I might use independant methods of Naive Bayes:

require "naive_bayes"

# Create independant perimeters.
colors = NaiveBayes.new(:black,
                        :white,
                        :grey,
                        :red,
                        :green,
                        :blue
)

shapes = NaiveBayes.new(:line,
                        :square,
)

# Train on specific data.

## Colors
colors.train(:black, 'black', 'black')
colors.train(:white, 'white', 'white')
colors.train(:grey,   'grey',  'grey')
colors.train(:red,     'red',   'red')
colors.train(:green, 'green', 'green')
colors.train(:blue,   'blue',  'blue')

## Shapes
shapes.train(:line,   '+----------+', 'line')
shapes.train(:square, '+----------+
|          |
|          |
|          |
|          |
|          |
',                                  'square')


my_color = "red"
my_shape = "+-------------+"

this_color = colors.classify(*my_color)
this_shape = shapes.classify(*my_shape)

puts this_color
puts this_shape

However when it suggested refinements, it was using a ruby gem that I wasn't even using. More often than not it suggests Python when I'm using Ruby. This is really frusterating.