Language:    

   

LeoStatistic
 

productpicture
Unit Price:   USD 52.45 
Total:   USD 52.45
plus 0% sales tax/VAT to USD 52.45: USD 0.00
USD 52.45
Delivery:   License key
Platform:   Windows 95, Windows NT, Windows 2000, Windows XP x32, Windows 98, Windows ME, Windows Server 2003 x32
Version:   11.11
Language:   English
 
Download Demo  Add to Cart Buy now

LeoStatistic
software for data presentation and statistical analysis and prediction.

Screenshots:
click on picture to enlarge

Click to enlarge
Distribution

Click to enlarge
Distribution of two variables.

Click to enlarge
Approximation
(constructor style interface).

Click to enlarge
3D view.

Click to enlarge
DOW trend.

  • Free download:
    leostatistic.zip (622 KB)
    (leostatistic.exe, leostatistic.hlp, leostatistic.cnt, leostatisticsetup.exe, leouninstal.exe)

  • Operation system:
    Window 95/NT and higher.

  • Version: 8.2 last modification at Aug. 15, 2003.

  • Type: Commercial. Password protected. Check registration page for more information.

  • Features:
    • Data
      • Import data series from database (MS Access) table, text file, copy/paste from spreadsheet (MS Excel).
      • Operate with more then one data series at the same time.
      • Virtually unlimited* number of records to present and analyze.
      • Automatically recognize calendar mm/dd/yyyy data format transferring it into yyyy.yyyy to use for market (currencies, shares, commodities) time trend analyzing.
      • There is an option to modify data series **.
    • Data presentation and statistical analysis.
      • Histograms of variables distribution and comparison with Student and Poisson curves.
      • X-Y chart of two variable dependence and curve fitting with least square method (variety of user interfaces to chose from), signals revealing and user defined formula. Automatically best polynomial and Fourie approximation.
      • Three dimensional in perspective projection presentation of variable as function of two arguments and distribution of two variables.
      • Color map style presentation of function of two arguments and distribution of two variables.
      • Quasi plane and parabolic least square approximation in multidimensional space of arguments (including case of two arguments where the fitting surfaces are visualized.
      • Nearest neighbors method of estimation value in multidimensional space.
      • A matrix of correlation coefficients of active variables.
    • Results.
      • Data and main statistical results is displayed on the screen in report-ready form. Statistical formulas and coefficients are shown in result panel.
      • User has ability to copy/paste screen chart as a bitmap picture into MS Office applications as well a content of result panel in form of a table.
      • There is an option to instantly create a MS Word document with a picture and results reflected to the content of chart screen and content result control.
      • A calculation of value for the given set of arguments for active statistical scheme when available (few logical exclusions).
      • A modeling of the array of arguments is defined in text file on the base of active statistical scheme with the option to sort results and save in text file. This option can be used for prediction of data behavior including selection of most perspective candidates (consumers).
    • Interface.
      • Transparent and self-explanatory user interface conveniently located at control panel of application.
      • At any moment a user via at most three clicks is able to reach any control to edit a statistical schemes or presentation options.
Screenshots:
click on picture to enlarge

Click to enlarge
Signals.

Click to enlarge
Near neighbor method.

Click to enlarge
Harmonic analysis.

Click to enlarge
Free format formula.

Click to enlarge
Crystal growth rate.

*Programmable limitation on the number of records is 232. Another limit of the software is time consumption and depends on the available memory and processor speed of the computer. For different statistical schemes times of calculations could be remarkable different. In general one can say that for less then thousand records any statistical schemes will perform practically instantly.

**The necessity to scaling in the data could appear for example if during an operation of curve fitting arguments are too large. There are common difficulties that during matrix inversion for example the precision of calculation with 2^64 bits presentation is not enough for the more or less sophisticated formulas. Very often meanwhile a simple normalization of initial data can solve the problem.
There are two reasonable solutions from developer's point of view: auto normalization of data and creation the option for user to do it himself. In this program the second solution is implemented.




 

Digital River GmbH (share-it!) is the seller of the products in your shopping cart and your contractual partner.

Copyright © 2010 share-it!