From 4ecbe8a87f9f209cfcf6d62ed226464fd240fab8 Mon Sep 17 00:00:00 2001
From: AJ Rossini <blindglobe@gmail.com>
Date: Sun, 9 Mar 2008 09:45:02 +0100
Subject: [PATCH] model thoughts and protos.

---
 model-fit.lisp | 100 +++++++++++++++++++++++++++++++++++++++++++++++++++++++++
 model.lisp     |   5 ++-
 2 files changed, 104 insertions(+), 1 deletion(-)
 create mode 100644 model-fit.lisp

diff --git a/model-fit.lisp b/model-fit.lisp
new file mode 100644
index 0000000..d0cc7fd
--- /dev/null
+++ b/model-fit.lisp
@@ -0,0 +1,100 @@
+;;; -*- mode: lisp -*-
+;;; Copyright (c) 2005--2007, by A.J. Rossini <blindglobe@gmail.com>
+;;; See COPYRIGHT file for any additional restrictions (BSD license).
+;;; Since 1991, ANSI was finally finished.  Edited for ANSI Common Lisp. 
+
+;;; File:       model-fit.lisp
+;;; Author:     AJ Rossini <blindglobe@gmail.com>
+;;; Copyright:  (c)2007, AJ Rossini.  BSD, LLGPL, or GPLv2, depending
+;;;             on how it arrives.
+;;; Purpose:    models as a data summarization tools.
+;;; Time-stamp: <2006-05-19 12:33:41 rossini> 
+;;; Creation:   <2006-05-17 21:34:07 rossini> 
+
+;;; What is this talk of 'release'? Klingons do not make software
+;;; 'releases'.  Our software 'escapes', leaving a bloody trail of
+;;; designers and quality assurance people in its wake.
+
+;;; The model class is used to map a partial or complete model
+;;; specification (partial specification made complete by assumptions
+;;; inherent in the type of model used, i.e. linear regression assumes
+;;; mean-zero and homoskedasticity) and the model-fit class is used to
+;;; complete the model specification into how the model instance and
+;;; type will then get computed.  The approach tken is that we
+;;; definitiely want to be able to see how we can explicitly try to
+;;; characterize the work that we are trying to infere about the data.
+
+(in-package :cl-user)
+
+(defpackage :lisp-stat-model-fit
+  (:documentation "Model fitting theory.")
+  (:nicknames :ls-model-fit)
+  (:use :common-lisp
+	:lisp-stat-config
+	:lisp-stat-object-system
+	:lisp-stat-types
+	:lisp-stat-compound-data
+	:lisp-stat-matrix
+	:lisp-stat-linalg)
+  (:shadowing-import-from :lisp-stat-object-system
+			  slot-value call-method call-next-method)
+  (:export fit))
+
+(in-package :lisp-stat-model)
+
+(defclass fit ()
+  ((criteria-functin-name :initform nil
+	 :initarg :name
+	 :accessor name
+	 :reader model-name)
+   (criteria-function :initform nil
+	 :initarg :formula
+	 :accessor function
+	 :reader fit-criteria)
+   (parameter-vars :initform nil
+		   :initarg :parameter-vars
+		   :accessor param-vars
+		   :reader model-formula)
+   (data-vars :initform nil
+	      :initarg :data-vars
+	      :accessor data-vars
+	      :reader model-formula)
+   (fixed-vars :initform nil
+	       :initarg :fixed-vars
+	       :accessor fixed-vars
+	       :reader model-formula)
+
+  (:documentation "Mathematical Model Fit approach"))
+
+(defclass optimization (model-fit))
+
+
+(defclass least-squares (optimization))
+(defclass weighted-least-squares (least-squares))
+
+(defclass maximum-likelihood (optimization))
+(defclass minimax (optimization))
+(defclass maximin (optimization))
+
+(defclass minimum-entropy (optimization))
+(defclass lq-norm (optimization))
+
+(defclass root-finding (model-fit))
+
+(defclass method-of-moments (root-finding))
+(defclass marginal-models (method-of-moments))
+(defclass gee (marginal-models))
+(defclass gee2 (marginal-models))
+
+
+
+;;; How would this be used?
+
+(setf my-regr-model-1
+      (new 'least-squares  '(- y (+ (* beta1 x1) (* beta2 x2)))))
+
+;; and there should be an approach which could provide a mapping to this, i.e. 
+
+(regression-model (list y) (list x1 x2))
+
+;; could map to the above via macros.
diff --git a/model.lisp b/model.lisp
index 8c69d1d..85e90c8 100644
--- a/model.lisp
+++ b/model.lisp
@@ -5,7 +5,8 @@
 
 ;;; File:       model.lisp
 ;;; Author:     AJ Rossini <blindglobe@gmail.com>
-;;; Copyright:  (c)2007, AJ Rossini.  BSD, LLGPL, or GPLv2, depending on how it arrives. 
+;;; Copyright:  (c)2007, AJ Rossini.  BSD, LLGPL, or GPLv2, depending
+;;;             on how it arrives.
 ;;; Purpose:    models as a data summarization tools.
 ;;; Time-stamp: <2006-05-19 12:33:41 rossini> 
 ;;; Creation:   <2006-05-17 21:34:07 rossini> 
@@ -96,6 +97,8 @@
 (defclass nonlinear-linear-regression-model (statistical-model))
 
 
+
+
 ;;; garbage follows
 
 
-- 
2.11.4.GIT