GEOMETRY application  user's comments

03262020, 10:19 AM
Post: #1




GEOMETRY application  user's comments
I would like to report a few noticed problems in the GEOMETRY application.
1. I draw any geometric figure (e.g. square or circle). Then I fill the inside of the figure with red. OKAY. Now I want to calculate the figure area. I type "area" in the Geometry Numeric View window, and I want it to be displayed in the PLOT window. And now the red color changes immediately to dark. If I do not give a sign that the field value is displayed on the calculator screen, the color does not change. 2. Why the program doesn't draw an ellipse or hyperbola when we type in the Symb window. curve pattern e.g. ellipse (x ^ 2/25 + y ^ 2/16 = 1 or hyperbola (x ^ 2/38y ^ 2/15 = 1). This problem is not there when we want to draw a circle or a Parabola. There is another problem with the parabola. The parabola has short arms. This needs to be improved. 3. I draw an ellipse and introduce Poin On to the ellipse. Then I want to run the "is element" test. And this is where strange things start. When I move point B along the ellipse, the message once appears: is element (GB, GA): 1 and once is element (GB, GA): 0. Similarly with hyperbola. This is a mistake to fix. 

03262020, 12:19 PM
Post: #2




RE: GEOMETRY application  user's comments
is_element is a CAS command, it is intended to work with exact equations and points, for approx equations or points, it may fail because of rounding errors. Maybe increasing the value of Epsilon (e.g. to 1e6) in the 2nd page of the CAS settings will help is_element accepting that the approx point is on the curve.


03262020, 12:53 PM
Post: #3




RE: GEOMETRY application  user's comments
I found another way to solve this problem. A test that works well for an ellipse or hyperbola is; is_concyclic.
The "iselement" test command works very well for all functions f (x) and f (x, y) and for geometric figures. A great tool. I don't know why it just doesn't work for hyperbola and ellipses (handdrawn and according to the formula). It doesn't help to change Epsilon. This improvement can be put off as urgent for now. 

« Next Oldest  Next Newest »

User(s) browsing this thread: