This FAQ contains the answers to the most frequently asked questions of our customers. Please check it first to see if it contains the answer to your question or the solution to your problem. If you cannot find the required information, please contact our support staff. To help us serve you better, please give as much information as possible so we can reproduce your problems - which Cygron product, what operating system, etc. Well get back to you as quickly as possible.
How much effort is required to learn using DataScope?
Not too much, or at least significantly less then for the common high-end data mining
tools. We strive to make that easy for everyone. DataScope comes with a quick tour,
interactive tutorial, sample databases and projects, detailed user's manuals.
What is included in the software package?
The package contains the following items:
How much time does it take to setup DataScope?
Unlike many data mining and decision support applications, DataScope installs very quickly and easily. The install program guides you through the steps of the installation process, which only takes about five minutes.
What are the system requirements?
DataScope was designed and carefully tested under Windows 95/98/2000/NT4.0. For Windows 95 and Windows NT 4.0, Internet Explorer 4.01 or later and the latest Windows NT Service Pack should also be installed. The required performance of the computer depends on the amount of data, but, as a minimum, an IBM PC or compatible with a 300 MHz or faster Pentium-class processor and at least 64 megabytes of RAM for Windows 9x or 128 megabytes of RAM for Windows NT/2000 is suggested. Also a minimum screen resolution of 1024x768 and color depth of 16-bit is strongly recommended to efficiently use the visualization techniques.
What are dongles?
Cygron uses Aladdin HASP dongles to copy protect its products. In order to run DataScope the user must plug in the HASP dongle key into the appropriate port of its PC. The dongles can be either USB port or Parallel port compatible.
What is the difference between the Local HASP and Net HASP dongles?
Local HASP dongles are for individual use. Each PC running DataScope requires a Local HASP dongle plugged in into one of its ports. Net HASP dongles can serve a network of DataScope users. Each PC running DataScope access and verifies the dongle plugged in into one of the PCs on the same network. Obviously client PCs must be connected to the same network as the PC with the dongle plugged in.
What kind of documentation is provided with DataScope?
Several information packages is included with various detail levels. Starting from the least detailed, they are:
How can I start working with a project in DataScope?
You can open any previously saved project file by using the Open command on the File menu. If you want to work with a new data table and create a new project file, you can do it by choosing the New command on the File menu and choose a method to specify your data (edit, import).
A dialog box button or menu command is disabled. Why?
Move your mouse over the disabled item, and hold down the left mouse button. A tooltip will explain the reason of the item being disabled.
Can I export my analysis results?
Yes, DataScope can print the data, charts, comments and more. You can also export this information to HTML and copy charts and data to the Clipboard. There is a possibility of exporting subsets of data into XML or a database via Microsoft ODBC.
What types of databases can DataScope analyze?
DataScope can access many popular database types through Microsoft ODBC, for example, FoxPro, MS Excel, MS Access, dBase, Paradox, text files as well as MS SQL Server, Sybase or Oracle. Multi-table database import is available using the DataMap module. Special support is included for Microsoft Excel data, which makes import as easy as a simple copy-paste operation. Due to speed problems with the ODBC text driver, DataScope also has a proprietary import algorithm for text files. Additionally, users have the option of manually entering information directly into DataScope or import data through the Clipboard. An easy-to-use import wizard lets you choose the importing method and asks the related questions.
How can I analyze multiple SQL or Access tables?
You will need the DataMap module for this. It allows a systems administrator or sophisticated end-user to create template files that describe the structure of a relational database. The templates can be distributed to end-users, and they can import data without having knowledge about the underlying data structure.
Can I detect and correct data errors of the imported data table in DataScope?
You can not only correct errors in your imported project file, but also insert new records or fields into the data table with the Edit Data command on the Tools menu. If you want to correct a data error in your imported project file, just choose this command, double-click the cell in the appearing spreadsheet you want to change the value in and edit the appearing text.
Detecting data errors is possible using the Tools/Data Quality Report command. This command calculates global database characteristics (e.g. number of n/a data, minima, maxima, averages of columns or custom mathematical expressions). It's easy to detect data errors by analyzing the report created by this tool.
What is the maximum amount of data that DataScope can handle?
There is no hardcoded limit to the amount of information loaded into DataScope. The size of the database it can load depends on the hardware configuration of the PC it is running on and operating system limits. The most important aspect is the total of the available physical and virtual memory. The processor performance only affects the speed of performing the various tasks. If the maximum virtual memory that can be used by a single process (2GB on Win98 and 1.5GB on Win2000/NT) is provided then DataScope can load approximately a 1 million record and 70 column database, ±15% depending on the data type of columns.
For analyzing databases that cannot fit into the memory DataScope offers sophisticated sampling techniques during data import, which help reducing the amount of information to be loaded and still retain the correct representation of the whole database by the sample.
How can I open a new window for a field in DataScope?
Use the commands in the Chart menu. To open a default chart for a field, use Quick Open. To open default charts for several fields quickly, use Quick Open/Multiple. To open multidimensional charts, use the New Chart Wizard command. The Relation Finder can also open charts to show the relations it has discovered. Use the Relation Finder command to do this.
Is it difficult to make queries?
No, querying and filtering operations are available as simple mouse operations and do not require learning a command set and/or formula language. You can create conditions like 'cars cheaper than $20000' and connect these with logical operations to make more complex queries. You can do many things with query results: color or filter diagrams by query results, delete other records from the database, and so on.
How much time does it take to run a query and see the result?
With traditional data query systems, you need to issue a query and wait for the result. With DataScope, it is easy to modify queries on the fly and see the resulting data set immediately. Through special preprocessing algorithms, DataScope always responds to a request immediately. This makes experimental discovery much quicker and easier.
Can DataScope display more than 2 dimensions of data?
Most charts can display multiple (3-8) dimensions of data or aggregated values. For example, you can display a classic 3D column chart, but you can also choose to assign additional dimensions to the height, width and color of each column.
How can I analyze a subset of records with DataScope?
If you want to analyze only a subset of the records imported to DataScope, you can make selections. To make selections, use the commands on the Condition menu to create their building blocks and then Selection menu to select the logical operation you wish to use to connect these blocks. A selection behaves like a discrete field and whenever you can use a discrete field (displaying, filtering, coloring), you can also use a selection. You can also choose to delete all records from the project that are not part of the current selection. To do this, create a selection, and choose the File/New command.
How can I set measurements and number format for columns?
If you want to set measurements for a numeric column, choose the Properties command from the Column menu. By choosing this command, you can also set the number of decimal digits you want to be used for the numeric values.
What is the use and function of the distribution chart?
This chart helps you to use subjective judgments instead of numbers. The percent value assigned to each record value shows the percentage of the records that have a lower value for the displayed field. For example, if the price $20,000 has a percent value of 12% in a car database, it means that 12% of the cars are cheaper than $20,000, so this is a relatively low price in this data set.
What types of algorithms does DataScope offer?
DataScope features the following algorithms to segment the data and to build models:
Clustering Methods: Kohonen, Shepherd, Fuzzy C-Means
In addition DataScope features a sophisticated Multi-Criteria Decision Making (MCDM)
algorithm, which provides ranking of decision alternatives based on weighted user
preferences and a relationship finder algorithm for automatic correlation detection
between all variations of numeric column pairs and triples and discrete column pairs.
Can I transform the data or create new data columns in DataScope?
Yes, DataScope can transform data using the following methods:
New columns can also be created from the results of the clustering or predictive data mining algorithms or ranking results of the Decision Support algorithm. These calculated columns represent results of complex calculations performed on other columns rather then their transformations.
Calculated columns can be based on other calculated columns, so they are hierarchically depend on each other. A very important feature of DataScope that all calculated fields are stored by their method of creation and are always recalculated if the underlying data or formula changes, and subsequently all related charts are updated to display the new information.
How the Decision Support Tool helps me with my problem?
The DataScope Decision Support Tool can rank decision alternatives based on your decision criteria. As a result, every alternative will have a preference value (and a rank based on that preference) assigned to it. This preference value and rank represents how much the given alternative suites your needs.
To start, you must define your decision criteria, that is, select the data columns, which store relevant attributes of the alternatives. Then continue by defining how your preference changes as the values within these columns change (the higher is better the lower is worse, or the opposite). Then you should compare these criteria, which is more and which is less important. This is done by assigning weights to criteria. You can use either numbers or natural language expressions (is much more important, slightly more important, etc.) in doing that. You can also group the comparable criteria so that only those within the same group should be compared with each other and at a higher level the groups are compared to each other. At the end of this process you will have a "criterion hierarchy" ready to rank your alternatives. As this "decision hierarchy" is built the rest is done by the MCDM algorithm, which calculates the preference values and some other attributes for each alternative. The higher the alternative's preference value is, the better it suites your needs. Note, that in certain circumstances it may happen that even highest preference value is low, meaning that none of the alternatives suites your needs really well.
The results are displayed on a special Decision Support chart, which allows you to evaluate the results in detail, compare two or more alternatives to each other at the level of their single attributes and won ranks.