I've rewritten an earlier application which calculates all the possible scores of probability, X², Student-t and F-distributions from the input and vice-versa. Also it shows a nice graphical view of the above, below, in between and outside probabilities of a give z-score.
I realise this is only for the statistical geeks like me, but I find it rather usefull in quickly computing the wanted scores.
Have fun
Any feedback is greatly appreciated
thnx in advance
Marc
Last edited by Stellaferox : 10-02-2009 at 07:50 PM.
Reason: forgotten attachment
Your program is similar to my Probability Calculator (see page two of this forum). Mine was designed to run on a WM cell phone, hence the popup number pad.
Two questions about your program: (1) when going from p(z) to z I'm getting the wrong answer. For example, for alpha = .05 (1-tailed) the program returns 2.271 -- the correct answer, of course, is 1.645. Also, the program will not allow me to enter an alpha value for t, so I cannot lookup the value of t associated with a given alpha, df.
You might also be interested in my binomial calculator on the last page of this forum.
I adjusted the program according to your remarks. Thanks for your input. Anything is welcome. I uploaded the new program in my original post.
Your program was actually the push to update my old Turbo-Pascal and Delphi programs!
Thank you for that. At the moment I am working on several programs, also one on statistics; more specific, a program that creates 2 huge populations and then let you choose of which population you want to take samples. It then computes the data according.
If you encounter more problems, please let me know.
Marc
I tried running your updated code (on the desktop), but now I receive the error message below.
I have a few Stat programs I may convert sometime, but there seems to be little interest so I have them on the back burner. My main focus right now is a GPS app, and then I may finish off my Let It Ride game.
A couple of more items. I hope you don't think I am being hypercritical, I'm just trying to help.
Why does the button in the upper right-hand corner of the main form now display "$" instead of "Calc"?
For the t, F, and Chi-square distributions, the random variable textbox is bottom of the group, and the alpha box is at the top. For the Z distribution, the order is reversed. This is confusing, but easy to fix.
I like the graph page a lot. It could benefit from a "calc" or "graph" button. Currently, if you enter a new value the only way to redraw the graph is to check another radio button and then recheck the one of interest.
You might consider having the last calculated values from the main form automatically entered into the appropriate places on the graph form. Then, the user can simply switch to the graph page and see the results.
Quite frankly, I appreciate your input very much because not a lot of people know or care enough about the subject to give feedback, so please continue to do so!
To address your comment:
- I changed it into a "$" because there wasn't any more space left to display "Calc". I thought about showing it in a pull-down menu but I wanted quick access. I guess if you rearrange the buttons a bit tighter, more space is left to display a proper button.
- About the order of the boxes: you're right. It is because I regard the z-score as basic to the Probability-scores. It is of course no problem at all to reverse the boxes and the cA.Checked line in the sub.
- About the graph-page: right again. I implemented this Graph page from an earlier written (stand-alone) program to calculate these scores. You don't have to check another radiobutton first, just rechecking the wanted button will do, but I agree that a "Calc" button here would suit in fine. I will do so.
- About you last remark: that is already on my to-do list. I also want to make a faster redrawing routine. This "Floodfill"routine is from earlier days when no real Floodfill-routine was available. On the desktop it works fast enough but on a pocket pc the routine is to slow.
The problem with most of my programs (or is it really my problem?) is that the moment functionality works I don't care very much anymore about presentation and displays. It is good therapy to make an extra effort for this, so thnx agian. In the following days I willl adjust the lines but feel free to adjust anything you'd like yourself. At the moment I am working on a program which shows the exact time of the sun coming up and going down world-wide, another stat-program I mentioned earlier and a game.
You are speeking of a Floodfill routine you are using.
Do you know that there is a Floodfill routine in agraham's ImageLibEx ?
I expect this one beeing much faster than one written in B4PPC.
Quick Statistics Application
Linear Mode
