FAME Time Series Database
|
From the publisher website or publicity FAME Time Series Database The FAME Time Series Database (Forecasting, Analysis and Modeling Environment) provides a robust solution for storing and managing high-volume time series data, defined as information collected at recurring intervals over a specific calendar period. Designed with a deep understanding of the unique and sophisticated properties of a time series, the FAME database is considered by many to be the de facto standard for the storage and management of such data. Currently used around the world by leading institutions in the financial, energy, and public sectors, FAME is an established, reliable solution offering a complete set of tools and APIs to quickly and easily integrate with downstream systems and custom applications. Unique "Time-Intelligent" Architecture FAME structures are specifically designed for time series data storage and manipulation. The FAME database manager maintains data on contiguous areas of the disk for the fastest possible retrieval of historical data. Unlike relational databases, the FAME database structure understands the concept of time. Most FAME database items have a common element – they relate to time. Instead of maintaining a date or time association in every database, FAME's foundation is a true calendar, interrelating every database point. The database has an implicit understanding of the period of time, enabling it to recognise months of varying lengths, business days, holidays and leap years. For example, FAME's underlying calendar can determine which years include 52 Fridays rather than 53, which July months have five Tuesdays instead of four, and accommodates interest rate calculations using either a 360- or 365-day basis. It's easy to understand how this understanding is critical to the fundamental analysis of economic and financial data. Within the FAME Time Series Database, information is stored at its natural frequency, rather than forcing disparate data types to conform to a single structure or model. Each value is associated with the actual calendar date on which it was reported. In a relational database, weekly data would be captured into 52 periods per year, even when some years have 53 observations. Within FAME, comparisons between items with different frequencies are handled by a seamless conversion to any specified working frequency. Special Time Series Attributes Time-series data has special characteristics that are part of the fundamental design of how the FAME Time Series Database assigns attributes to each time series, a concept unknown in relational database systems. These attributes define each data type: end-of-period values (inventory), beginning-of-period values (opening prices), summed values (sales), annualised values (housing starts) and highs or lows (equity prices). FAME makes intelligent use of these attributes when converting data from one frequency to another. Time series can have user-defined attributes in addition to those defined by FAME. Assigning user-defined attributes to a time series enables customised classification and screening. e.g. custom attributes like country and currency could be used to locate all time series relating to France denominated in U.S. dollars. Bond attributes such as issue-type, maturity, and coupon might be instilled, enabling the filtering of the database for specific maturities. The range of applications is virtually limitless. Highly Tuned and Optimized Database Structure The FAME Time Series Database is optimized for rapid retrieval of time-series data, employing a sophisticated indexing system, contiguous data storage, and dynamic caching intelligence to speed the retrieval of time-series data. Each time series is stored as a contiguous block, allowing all or part of the series to be scanned with a single disk read. FAME's database manager maintains data continuity by intelligently allocating disk space during the update procedure. Multi-level, read-write caching at the object, index and storage-block levels minimise disk access retrieval times. Flexible Data Storage Each time series in a FAME database can span different data ranges. For example, a single FAME database can mix annual Net Exports data with quarterly corporate earnings and daily pricing data.
Database Access - APIs
Tools and Utilities FAME Emulator - The FAME Emulator works with accessPoint, allowing FAME 4GL, TimeIQ, C HLI and queryPoint applications that normally talk to an FRDB server to pull data from an accessPoint server.
|
Keywords
Other tools that may interest you by area:-
Vendor: Sungard |
|