A wrapper around fityk (fityk.nieto.pl/) to handle fitting a curve to X+Y data, creating confidence intervals, and projecting up to a ceiling.
You must have a working installation of cfityk in your path for this library to work. I have v1.3.0.
The primary use of this library is to take a data array of X+Y points and return to you a trend line, confidence intervals, a best-guess shape and an r-squared value.
require 'curve_fit' cf = CurveFit.new cf.fit( [ [ 0, 1000.0 ], [ 1, 2003.0 ], [ 2, 3010.0 ], [ 3, 4084.0 ], [ 4, 5012.0 ], [ 5, 6075.0 ] ] )
Returns:
{
:data => [
[0, 1000.0],
[1, 2003.0],
[2, 3010.0],
[3, 4084.0],
[4, 5012.0],
[5, 6075.0]
],
:trend => [
[0, 997.27],
[1, 2010.57],
[2, 3023.87],
[3, 4037.1699999999996],
[4, 5050.469999999999],
[5, 6063.77]
],
:top_confidence => [
[0, 1010.5636999999999],
[1, 2030.03089],
[2, 3049.49808],
[3, 4068.9652699999997],
[4, 5088.43246],
[5, 6107.899649999999]
],
:bottom_confidence => [
[0, 983.9763],
[1, 1991.1091099999999],
[2, 2998.24192],
[3, 4005.3747299999995],
[4, 5012.50754],
[5, 6019.64035]
],
:ceiling=>[],
:r_squared=>0.999774,
:guess=>"Linear"
}
You can pass a second argument, which will be the ceiling to project up to. In that case, the trend, top and bottom confidence data will be extended out until the Y value is >= the ceiling. Additionally, the ceiling line will be populated with the ceiling value for each value of X.
cf.fit( [ [ 0, 1000.0 ], [ 1, 2003.0 ], [ 2, 3010.0 ], [ 3, 4084.0 ], [ 4, 5012.0 ], [ 5, 6075.0 ] ], 10000 )
Would extrapolate up to 10000.
Currently, this library only supports linear, quadratic, cubic, polynomial4, polynomial5, and polynomial6 fits. By default, it tries all, in that order. You can specify the fits to guess manually if you want to narrow things down:
cf.fit( [ [ 0, 1000.0 ], [ 1, 2003.0 ], [ 2, 3010.0 ], [ 3, 4084.0 ], [ 4, 5012.0 ], [ 5, 6075.0 ] ], 10000, [ "Quadratic", "Linear" ] )
Finally, you can pass a block to manipulate the value of X in the result set. This is often used if your data is a time series - under the hood, we convert the values of X to the position in the array (0..n). This block lets you get the original data back out, and support generating new correct values as you extrapolate to a ceiling.
For example, lets assume we’re daily from 2010/1/4.
cf.fit( [ [ Time.utc("2010", "1", "4").to_i, 1000.0 ], [ Time.utc("2010", "1", "5").to_i, 2003.0 ], [ Time.utc("2010", "1", "6").to_i, 3010.0 ], [ Time.utc("2010", "1", "7").to_i, 4084.0 ], [ Time.utc("2010", "1", "8").to_i, 5012.0 ], [ Time.utc("2010", "1", "9").to_i, 6075.0 ] ], 10000 ) do |x| if x == 0 Time.utc("2011", "1", "4").to_i else (Time.utc("2011", "1", "4") + ( 60 * 60 * 24 * 1 )) end end
Will result in the data set, and extrapolated data, having seconds since the epoch for X values.
-
Check out the latest master to make sure the feature hasn’t been implemented or the bug hasn’t been fixed yet
-
Check out the issue tracker to make sure someone already hasn’t requested it and/or contributed it
-
Fork the project
-
Start a feature/bugfix branch
-
Commit and push until you are happy with your contribution
-
Make sure to add tests for it. This is important so I don’t break it in a future version unintentionally.
-
Please try not to mess with the Rakefile, version, or history. If you want to have your own version, or is otherwise necessary, that is fine, but please isolate to its own commit so I can cherry-pick around it.
Copyright 2010 Opscode, Inc.
Licensed under the Apache License, Version 2.0 (the “License”); you may not use this file except in compliance with the License. You may obtain a copy of the License at
http://www.apache.org/licenses/LICENSE-2.0
Unless required by applicable law or agreed to in writing, software distributed under the License is distributed on an “AS IS” BASIS, WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. See the License for the specific language governing permissions and limitations under the License.
See LICENSE file for complete license details.