~rgrjr/rgr-hacks

c407f7b804093bb100955356b997ad4d59908a83 — Bob Rogers 2 months ago 6082cd0
Also predict the pumpernickel week after next

* rgr-pumpernickel.el:
   + (rgr-pump--make-prediction):  Pass in a title, do our own
     probability normalization. and fix the fill-paragraph so it works.
   + (rgr-alist-convolution):  Quick and dirty.
   + (rgr-show-pumpernickel-intervals):  Predict the next pumpernickel
     week, and the one after.
1 files changed, 35 insertions(+), 10 deletions(-)

M rgr-pumpernickel.el
M rgr-pumpernickel.el => rgr-pumpernickel.el +35 -10
@@ 53,7 53,8 @@
    (dolist (pair hist)
      (let ((days (car pair)) (count (cdr pair)))
	(setf (aref histogram (floor days 7)) count)))
    ;; Allocate three characters to each week.
    ;; Make the drawing, allocating three characters to each week.
    (insert "\n")
    (cl-do ((height 9 (1- height))
	    (str (make-string (* (length histogram) 3) ?\ )
		 (make-string (* (length histogram) 3) ?\ )))


@@ 69,13 70,17 @@
	(insert (format "%3d" idx))))
    (insert "  weeks\n")))

(defun rgr-pump--make-prediction (last-day n hist)
(defun rgr-pump--make-prediction (prediction-title last-day hist)
  ;; last-day is the rgr-date-to-day of the most recent pumpernickel week for
  ;; basing predictions, hist is an alist of (days . count) sorted in descending
  ;; order of count, and n is the total count.  Show all predicted dates with
  ;; more than 20% of the total probability.
  (insert "\nNext week is estimated to start ")
  (let ((need-comma nil))
  ;; order of count.  Show all predicted dates with more than 20% of the total
  ;; probability.
  (insert "\n" prediction-title " is estimated to start ")
  (let ((need-comma nil)
	(n 0))
    (dolist (pair hist)
      (let ((count (cdr pair)))
	(cl-incf n count)))
    (dolist (pair hist)
      (let ((prob (/ (* (cdr pair) 100.0) n)))
	(when (>= prob 20)


@@ 85,7 90,23 @@
			  prob))
	  (setq need-comma t)))))
  (insert ".\n")
  (fill-paragraph))
  ;; Now pretty it up a bit.
  (save-excursion
    (forward-line -1)
    (fill-paragraph)))

(defun rgr-alist-convolution (alist)
  ;; Given an alist of (key . count), produce its self-convolution.
  (let ((result nil))
    (dolist (pair1 alist)
      (dolist (pair2 alist)
	(let* ((key (+ (car pair1) (car pair2)))
	       (count (* (cdr pair1) (cdr pair2)))
	       (result-pair (assoc key result)))
	  (if result-pair
	      (cl-incf (cdr result-pair) count)
	    (push (cons key count) result)))))
    (sort result #'(lambda (a b) (> (cdr a) (cdr b))))))

;;;###autoload
(defun rgr-show-pumpernickel-intervals ()


@@ 126,10 147,14 @@ The dominant intervals correspond (so far) to 6 or 5 weeks."
	      (cl-incf n count)
	      (push (cons i count) hist))))
	(setq hist (sort hist #'(lambda (a b) (> (cdr a) (cdr b)))))
	;; (insert (prin1-to-string hist))
	(insert "\n")
	(rgr-pump--draw-histogram hist)
	(rgr-pump--make-prediction last-day n hist)))))
	(rgr-pump--make-prediction "Next pumpernickel week" last-day hist)
	;; Now do the two-interval prediction.
	(let ((twice-hist (rgr-alist-convolution hist)))
	  ;; (insert (prin1-to-string twice-hist))
	  (rgr-pump--draw-histogram twice-hist)
	  (rgr-pump--make-prediction "The following pumpernickel week"
				     last-day twice-hist))))))

;; (/ (rgr-days-between "2023 Dec 13" "2023 Oct 18") 7) 8