de Dimension 5, Ltd
Miner3D es un software de visualización para el análisis y la exploración avanzada de datos multidimensional. Una intuitiva, moderna y muy potente interfaz de usuario le ayuda a realizar su trabajo diario de un modo mucho más eficaz.
Miner3D is a data visualization software for multidimensional exploratory
data analysis. Advanced, powerful and intuitive user interface helps a user to
work effectively without extensive training.
Models: Scatter, Bar, Line, Histograms, Tiles, Heat Maps
Integrated model builders automatically create charts for currently available
data. Miner3D includes support for making several basic 3D and 2D charts that
can be quickly and easily customized and even converted to a totally different
graph type. High level of flexibility enables a user to generate virtually
unlimited amounts of possible combinations of two or three dimensional charts.
Selector is a powerful data selection tool that enables a user to easily "Slice
and Dice" your data and display it in a variety of interactive views. It
provides a visual and interactive query making and helps to reveal patterns and
clusters in data and their layouts and decompositions in data models. This tool,
in combination with the Statistics calculator, plays a central role in Miner3D
software enabling thorough data analysis with a high level of interactivity and
The Miner3D Selector changes the way people are used to perform business
intelligence by creating a fast and convenient access to building your own
interactive OLAP Cubes. In many types of tasks it perfectly eliminates the need
for a traditional query & reporting tools, complicated online analytical
processing or for any other separate "heavy-analytics" technology.
Statistics calculator computes selected results and summaries on data subset,
which is currently selected by Selector. This tool automatically performs
user-defined calculations and helps a user to quickly measure and analyze data.
It provides additional information that may be easily combined with information
discovered through pure visualization.
Statistics with the support from Selector provides a user the same
functionality as running CUBE, ROLLUP, COMPUTE commands on complicated and
hard-to-use SQL technologies. For most of users, the use of Miner3D?s
interactive and visual approach is a perfect replacement for complicated OLAP,
Cubes and SQL queries.
Interactivity, ease-of-use and the presence of query and visualization
context, combined with statistic data, allow a user to quickly and easily find
data subsets that produce requested results.
K-Means Clustering (Miner3D Enterprise only)
K-Means Clustering and K-Means Data Reduction will give you more options to
process large data sets. The popular method can be used either to cluster data
sets visually in 3D or for row reduction ad compression of large data sets.
Our implementation of K-Means uses a high-performance proprietary algorithm
based on the filtering algorithm and multidimensional binary search trees.
Principal Components Analysis (PCA)
Principal Component Analysis will give you more options to process
high-dimensionality data sets, consisting from many columns. These methods allow
you to prepare data subsets that are more suitable for data visualization and
increase the chance to see good results from your analysis.
PCA extracts the key information and throws away the rest. Improved PCA will
assist you to process high dimensionality data sets. PCA in Miner3D is
implemented via eigen decomposition of the covariance matrix.
Miner3D allows you also to create video files where your data analyses are
recorded step by step. You can have saved in live video shots a process of
refining data model, movements, zooming, rotations and also visual querying in
The movie files can be played in Windows Media Player® or other video player
software. You also can insert video files to multimedia presentations, or
emailed to your colleagues or team members.
Data Row Reduction
Row Reduction helps to manage the number of data points and keeps the
interactivity and the overall feedback of a data model at an acceptable level.
Available data sampling methods: - K-means Clustering: rows clustering using
- Random selection: select a percentage or an exact number of rows to be
- Uniform selection: select a step value (every N-th row), and/or range to
- Microsoft SQL Server®, Microsoft Access®, Oracle®, IBM DB2®, MySQL® or
other SQL database engines and servers (Miner3D Enterprise only)
- Microsoft Excel 2003, XP, 2000
- Windows Clipboard
- CSV files (comma separated values)
- TXT files (tab delimited data)
- DBF files
- Miner3D Model files (M3D)
- Data files (CSV, TXT)
- Clipboard Data and images
- High resolution presentation-quality image files in several bitmap formats
(BMP, JPEG, TIF, TGA, PCX, RGB)
- Movie files containing digital video record