Auguri

Auguri is a statistical, data exploration, analysis and forecasting software with emphasis in nonlinear dynamical methods that provides the data manipulation tools required in predictive data mining and advanced analytics.

Auguri Features

Data Import and Export
ANSI Data Import and Export.
Binary Data Import and Export.
Copy and Paste.
Drag and Drop.

Data Editing, Formatting and Printing
Automated data population
Editing, formatting, and printing
Mixed data type support: Date, Text, Number, and Formula
Find and Replace Operations
Cut, Copy, and Paste
Drag and Drop
Sorting

Charting
Multiple chart types: Line, Bar, Point, Surface, Contour, and so forth
One-, two-, three-, and four-dimensional charts
Conversion among compatible chart types
Chart Animation
Chart formatting, printing, saving and exporting

Data Model Definition
Easy model definition and specification
Automatic embedding
Automatic assignment of values for dates and text
Unlimited number of concurrent model solutions
Multivariate Models of up to 8,192 vectors of 1,024 elements each and 4,194,303 instances

Math Operations Between Series
Addition
Subtraction
Multiplication
Division
Logical AND
Logical OR
Logical XOR

Analysis Tools
One- and two-factor Analysis of Variance.
Auto and cross Average Mutual Information.
Auto and cross Covariance and Correlation Functions.
Chi-Square Test for one Population Variance.
Descriptive Statistics:
Mean
Median
Mode
Geometric Mean
Harmonic Mean
Mean Deviation
Root Mean Square
Variance
Standard Deviation
Sample Variance
Sample Standard Deviation
Standard Error
Skewness
Standard Error of Skewness
Kurtosis
Standard Error of Kurtosis
Count
Sum
Range
Minimum
Maximum
Confidence Interval
Additional Modes
False Nearest Neighbors
Frequency Domain Correlation
Generalized Dimensions under different numerical methods:
Ellner
Grassberger-Procaccia
Takens-Theiler
and others
Histograms:
Pareto
Natural
Uniformly Binned
Gaussian-Binned
IID Tests:
Box-Pierce
Difference-sign
Ljung-Box
McLeod-Li
Rank
Turning Point
Maximal Lyapunov Exponent:
Kantz
Rosenstein
One-Sample Tests for Means:
t-Test
z-Test
Poincare Surface of Section
Power Spectrum Estimation:
Periodogram
Averaged
Windowed
Maximum Entropy
Recurrence Analysis
Running Statistics:
Progressive or windowed
Mean
Root Mean Square
Variance
Mean Deviation
Standard Error
Standard Deviation
Simultaneous Solution of Linear Equations
Space Time Separation Plot
State Space Visualization (Phase Portraits) in up to 4-dimensions
Two-Sample F-Test for Variances
Two-Sample Tests for Means:
t-Test
z-Test

Operation Tools
User-defined functions with functional equation parser
Window Functions:
Barlett
Blackman
Blackman-Harris
Dolph-Chebyshev
Half-Cycle Sine
Hamming
Hann
Kaiser
Parzen
Welch.
Difference and Summation of series.
Digital Filter Design:
Sinc Function
Remez Exchange
Frequency Custom.
Embedding
Event Times and Times Event
Exponential Smoothing
Fourier Transforms for one or two dimensions:
Mixed Radix Real and Complex
Frequency Domain Convolution
Numerical Interpolation and Resampling for one- and multi-dimensional uniformly and arbitrarily spaced data.
Moving Average.
Data Normalization: Zero Mean One-Standard Deviation with optional scaling.
Numerical Differentiation for empirical data for uniformly and arbitrarily spaced data.
Numerical Integration for empirical data for uniformly and arbitrarily spaced data.
Polynomial Expansion.
Automated data population.
Random Number Generation in several distributions:
Uniform
Beta
Binomial
Chi-Square
Exponential
F-Distribution
Gamma
Gaussian
t-Distribution
others.
Data Sampling (see Random Number Generation for sampling distributions).
Savitzky-Golay for uniform and arbitrarily spaced data
Data Scaling
Surrogate Data Generation:
Random Shuffle
Phase-Randomized
Gaussian Scaled
Fourier Shuffled
Iterated Amplitude Adjusted
Multi-Dimensional Fourier Transformed

Table and Data Operations
Forced Text Removal
Joining and Splitting
Row and Column order reversal
Text Value Classification
Transposition

Model Approximation Methods
Static Global Least Squares: Solution is based on a static data section
Dynamic Global Least Squares: Solution data section changes dynamically according to the prediction point
Global Multilayer Perceptrons: User Defined Feedforward Artificial Neural Networks
Averaged K-Nearest Neighbors
Weighted K-Nearest Neighbors
Local Least Squares
Local Multilayer Perceptrons
Editing of Solutions
Test Reports
Forecast Error Analysis
Interactive Tests and Simulation
Simultaneous Solution Comparison
Run Solutions against external data
Formatted account of test results
 
< Prev   Next >