CA2085681C - Indexing/compression scheme for supporting graphics and data selection - Google Patents

Abstract

A method for interactively selecting and displaying a distribution of all data fields in a selected database. The primary function of the methodology is to allow a user to view the data contained within a selected database in a graphical format. Additionally, the user has the capability of selecting specific sets of data for more detailed display as well as determining the effect of the selection on all remaining data fields which comprise the selected database. The first step in the process is the accessing of the particular database from the host system. Once the selected database has been accessed, a distribution matrix for each field in the selected database is constructed from the data contained within the selected database. The distribution matrix is then graphically displayed on any type of graphical display unit. From this initial graphic display, the user can make various selections for detailed graphical display and study and analyze the data structure and data content of the database.

Description

20~5681 INDEXING/COMPRESSION SCHEME FOR SUPPORTING GR~PHICS AND
D~TA SELECTION
:
The present invention relates to a general indexing and compr'ession scheme for supporting graphics and data selection, and more particularly, to a method for displaying '~
and selecting a distribution of all data fields in a selected database. The method of the present invention allows a user to choose fields in a particular database, build a distribution of the data in the fields, graphically view the data structure, and select specific sets of data to determine the effect of adding or subtracting data to the selected sets on the remaining data fields.
Over the past twenty years or so, advancement in the art of computerized information storage and retrieval has significantly expanded man's capability for efficiently accessing information. Todays computers with enhanced integrated circuit technology are capable of storing tremendous amounts of information, or to be more exact, data.
Accordingly, there is an ever increasing need for the development of systems and processes capable of managing the data and which permit the efficient utilization of the data.
The efficiency of computerized information storage and retrieval systems is directly related to how efficiently the database can be searched and how quickly records from the database can be retrieved. There are a plurality of 5 systemsand methods currently available which utllize various combinations of filing schemes, indexing schemes, and data compression schemes to enhance the efficiency of data manipulation; however, the efficient retrieval of data is really only the first step in the process. The display of 3~ the data in a way that is meaningful to the system user is a ~ ~
second and necessary step of the process, because no matter ' how fast one can access data, if the data retrieved is arbitrarily displayed on the screen or monitor, then the data is meaning]ess to the user. In addition to displaying the ;

2~85681 ' -2-1 data in a meaningful way, the system user should have the capability of selectively displaying certain detailed aspects of the data structure and determining the effect of adding or ;~
deleting data to the selected sets on the rem~ining data 5 fields.
The prior art references individually disclose systems and methods for the filing, indexing, compression, retrieval, and the display of data in a computer system; - -however, there appears to be no reference disclosing a system 10 or method for the filling, indexing, compression, retrieval, and display of data in one single embodiment wherein the -~
system user has the capability of selecting specific sets of data to determine the representative effect on the remaining data fields. The following references are representative 15 examples of state of the art systems and methods capable of performing some of the above listed functions, but not all.
In U.S. Patent No. 4,817,036, Millett et al.
dlscloses a computer system and method for database indexing and information retrieval. A number of keywords are selected 20 and each record of a database is searched to determine in which records each keyword appears. The central processing unit of the system then creates a vector of each keyword which identifies each record number of the database where the keyword appears and numerically sorts the record numbers. A
25 special bit processor next transforms each vector into a bit string that is identified by one of the keywords. The bit strings are returned to the central processing unit and -~
stored in secondary storage so as to form an index for the database. To retrieve information, one or more keywords are 30 lnput to the central processing unit. The input keywords are used by the central processing unit to identify the bit string for each keyword. The keywords may be logically ~oined using "AND", "OR", and/or "NOT" commands. Each bit ~tring retrieved from the index is then sent to the special '~
35 purpose bit processor, which combines the bit strings 2085681 ~

~.

-3- ;~
1 according to the particular command. The resultant bit string is trans~ormed by the bit processor into a vector which is returned to the central processing unit and which then is used to identify the individual records which contain 5 the combined keywords.
In IJ.S. Patent No. 4,961,139, Hong et al.
discloses a database management system for real-time applications. A real-time database provides the predictable, high speed data access required for on-line applications, 10 while providing flexible searching capabilities. The data retrieval routines include the option to "read-through-lock"
to access data in locked data tables, the capability to ~ ~
directly access the data using tuple identifiers, and the ~ -capability to directly access unformatted data from input 15 areas which contain blocks of unformatted data. The data updating routines include an option to omit index updating when updating data and an option to update data in a locked data table. Multiple indexes can be defined for a data table. Thus, high speed searches can be performed based on a 20 variety of data fields. The data storage and retrieval mechanlsms are independent and there are hash index tables ~;
that connect the multiple index keys to the data tables. ~ ~
The data table structure lncludes a column defined for ~;
storing tuple identifier strings. These tuple identifiers 25 can be used as pointers for chainlng to related data stored in other data tables. The database has relatively small programmable memory. There is a common structure for user data tables, index tables, and system data tables. The database includes a minimum number of routines wlth certain 30 routines providing multiple functionality.
In V.S. Patent Mo. 4,232,375, Paugstat et al.
discloses a data compression system and apparatus. The ~ -system compresses a binary data message generated by a digital input device. The compression process is based upon 35 the deletion of redundant information. The data compression '''~:' 20~5681 - ~

1 apparat~ls includes means for storing portions of a first data message generated by a terminal device as the result of a merchandising transaction performed during the time the CPU
is disabled, counter means for deleting all redundant data 5 characters of each data message, means for comparing preselected data characters of each succeeding data message with the corresponding data characters of the first data message stored in the terminal device for deleting those data characters when a comparison is found, and table look-up 10 means for selecting a start of record character in accordance wlth a data character representing the type of data transaction being processed, the start of record character indicating the start of the compressed data record in addition to the transaction type where there i5 an absence of 15 data in another portion of the compressed data record. -The first reference is representative of a group of references directed to systems and methods for the indexing and retrieval of data in a database. The second reference is representative of a group of references directed to the 20 creation of databases and database management systems. The third reference is representative of a group of references directed to systems and methods for the compression of data to avoid wasting memory. None of the cited references, however, specifically discloses a system which utilizes a 25 process wherein a distribution of data is built from the data contained ln a particular database and then graphically displayed with the capability to selectively add or delete data to the distribution and determine the effect of such addltion or deletion of data on the remaining data fields in 30 the distribution built from the data contained within the particular database.
The present invention is directed to a method for displaying and selecting a distribution of all data fields in a selected database. The method for displaying and selecting 35 of the present invention provides for the construction of a 2~5~ ~

1 distribution matrix for each field in a given record which allows analysis of the selected ~atabase to incorporate graphical,visualization of the data structure and content.
The primary function of the distribution is to allow the user 5 to view the data in a graphical format; however, the method also provides the user with the capabi]ity of selecting specific sets of data to be displayed and to determine the effect of the selection on all remainin~ data fields whlch comprise the selected database. The first step in the 10 process includes the step of accessing the particular database from a host system having at least one database resident thereon. The databases that can be accessed are the original databases in vendor format. The database is accessed in order to build the distribution from the data in :
15 the database. The next step in the process is to construct the distribution from the data contained in the database.
~ " . ,.
The construction of the distribution or distribution matrix is a multi-step process which prepares a representation of the data contained in the database for display and selection.
20 Once the construction of the distribution is complete, the display of the distribution for the selected data fields in the selected database is performed. At this point, the user can make "what if" selections and thereby select specific sets of data to determine the effect of the selection on all 25 of the data fields which comprise the selected database and ; ~' display a detailed distribution of the data based upon the step of selecting specific sets of data. '~
The method of the present invention provides a PC- ~ ~-based tool for access and analysis of information across 30 heterogeneous databases. It is intended to enhance costly ;
and inefficient database management systems and to further ' provide a powerful graphical user interface for comprehensive ' viewing of the data. The method incorporates a highly efficient matrix indexing technology which allows for rapid 35 access to the information contained within any of a number of : ~
'!.,, ' ' ' ' . . " " ; ' 2~85681 . ~

1 databases, and allows for interfacing to major database management systems such as IDMS or SQL/DS for mainframe computers, Oracle or Sybase for mid-range computers, and Paradox or Dbase for micro computers. The distribution 5 construction process utilizes processes which vary for each data type a]lowed in the various databases which provides the present invention with great versatility.
The present inventlon provides for a user friendly, completely graphical and menu driven software package which 10 provides a method for displaying and selecting a distribution ~ ' of all data fields in a selected database. A user of the software package can view the contents of the database in graphical format and make selections of specific sets of data currently displayed and then view the detailed selection and 15 its effect on the remaining data associated with the primary selection. The software package is simple to install, and only requires approximately 250K of memory. The software package is an order of magnitude faster than currently available query systems and is compatible with most commonly 20 used standard databases. The software package can also be utilized with non-standard databases. For non-standard -~
databases, an entry screen is used to define application generated record structures. The entry screen provides a series of questions enabling the system user to define the 25 application.
Figure l is high a level flow chart representing the software necessary to implement the method for displaying and selecting the distribution of all data fields in a selected database of;the present invention.
Figure 2 is a detailed flow chart representing the database access software of the present invention.
Figure 3 is a detailed flow chart representing the distribution construction software of the present invention.
Figure 4 is tabular representation of a sample of 35 database records and their corresponding field values.

2~8~681 1 Figure 5 is a tabular representation of BIN ranges for the field values illustrated in Figure 4.
,Figure 6 is a diagrammatic representation of a sample field header data structure utilized in the present 5 invention.
Figure 7 is a block diagram representation of the matrix work areas of the present inventlon. ~ -Figure 8 is a diagrammatic representation of the ~-distribution matrix for record number 1.
Figure 9 is a diagrammatic representation of the distribution matrix for record number 2.
Figure 10 is a diagrammatic representation of the distribution matrix for record number 3.
Figure ll is a diagrammatic representation of the ~
15 distribution matrix for record number 4. '~ -Figure 12 is a diagrammatic representation of the ;
distribution matrix for record number 5.
Figure 13 is a diagrammatic representation of the ~;
distribution matrix for record number 6.
Figure 14 is a dlagrammatic representation of the dlstribution matrix for record number 7.
Figure lS is a diagrammatic representation of the dlstribution matrix for record number 8. ., Figure 16 is a diagrammatic representation of the ~
2 distribution matrix for record number 9. ~-Figure 17 is a diagrammatic representation of the ~
distribution matrix for record number 10. ~-Figure 18 is a diagrammatic representation of the -distribution matrix for record number ll.
Figure l9 is a tabular representation of the final data content of the BIN work area. -Figure 20 is a tabular representation of the compressed final data content of the BIN work area.
Figure 21 is a detailed flow chart representing the 35 Elle opening soEtware oE the present in~entlon.

:

. ~

- 20~56~

Figure 22 is a detailed flow chart representing the data expansion software of the present invention.
,Figure 23 is a sample primary graphical display output of the present invention. ;~;
Figure 24 is a primary graphical display of the SAMPLE01 records output of the present invention.
Figure 25 is a detailed flow chart representing the secondary display software of the present invention.
Fiqure 26 is a sample secondary graphical display ~; -lO output of the present invention. ~ ~-Figure 27 is a first detailed flow chart representing the testiary display software of the present invention.
Figure 2~ is a second detailed flow chart ~ ~-15 representing the testiary display software of the present invention.
Figure 29 is a sample testiary graphical display and list box output of the present invention.
Figure 30 is a sample graphical display output 20 illustrating the selection process of the present invention.
Figure 31 is a sample graphical display output illustrating the deselection process of the present invention.
Figure 32 is a sample graphical display output 25 illustrating the ripple effect of the deselection process of the present invention.
Figure 33 is a flow chart representing the entire graphical interface software routine of the present invention.
The present invention is directed to a method for displaying and selecting a distribution of all data fields in a selected database. A database is an organized array of files where data files are fully integrated and the online access of data from these files is possible. A given 35 database file contains a number of fields grouped together e 20~81 g ..
1 to form a record of information. Each record is a complete set of information that can be used in the analysls of the -content of the database. The present invention provides for the creation of a distribution matrix for each field, in a 5 selected database which allows analysis of the database to incorporate graphical visualization of the data structure and content. The fields are processed to build a complete matrix -~
array of the field information. The structure of the matrix array shall support the process to display the data content 10 of the field. The primary function of this distribution is ~-to allow the user to view the data content of the file in a graphical format. However, the present invention also provides the user with the capability of selecting specific - ....:
sets of data to be displayed graphically and to determine the 1~ effect of the selection on all the remaining data fields -;~
which comprise the particular database if desired.
Referring to Figure 1, there is shown a high level flow chart representing the software necessary to implement the method for displaying and selecting the distribution of 20 all data fields in a selected database. The flow chart of Figure 1 illustrates the overall process for displaying and selecting the distribution of data fields, and as each element or block in the flow chart is described, reference to detailed flow charts and graphic figures are made to provide 25 detailed information for each step in the overall process, and to illustrate the process itself. The entire software package is collectively known as dbEXPRESS M
The entry point into the software routine is represented in the flow chart by element 10, which is simply -~
30 the start or initialization point of the program. Element ~ -100, which is the element directly following element 10, represents the software to implement the process of accessing a particular database from wherever the particular database is stored on the system. The databases may also be imported 35 from Local Area Networks (LAN's), communication ports and 2~5681 ~ ~

-10- .
1 from a variety of other areas. The databases that can be -accessed are the original database files in vendor format that are ~esident on the particular system. The present invention, dbEXPRESSTM, is a software tool for access and 5 analysis of information across heterogeneous databases such as, but not limited to, Paradox and Oracle for micro-computers, Ingres and Sybase for mid-range computers, and DB2 and IMS for mainframe computers. Element 100 also represents the software necessary to convert the particular database 10 file into dbEXPRESSTM operational, schema, and data files, from which the distribution is constructed.
Figure 2 i9 the detailed flow chart of the database access routine illustrated in Figure 1 as element 100.
Element 102 is the point of entry into the database accessing 15 routine. Element 104, which is the element directly following element 102, represents the software to implement the process step of retrieving the name of the particular database and the particular data tables within the particular database. The names are retrieved to access the data 20 contained therein. The retrieval process is similar to indicating a path as utilized in DOS in the sense that the path names provide a way in which to access particular files.
Element 106, which is the element directly following element 104, represents the software to implement the process step of 25 opening the particular database so that the data can be read from the particular database and into dbEXPRESSTM. In addition, element 106 represents the software to implement the process step of opening the internal setup or schema of dbEXPRESSTM so that the data from the database may be 30 accurately and efficiently transferred into dbEXPRESSTM.
Eiement 108, which is the element directly following element 106, represents the software to implement the process step of eliminating the database schema so that only the raw data remains. This software deletes the schema records for all 35 fields in the specified tables inithe database and exposes 2~5681 1 the data so that it may be easily read into dbEXPRESSTM
Element 110, which is the element directly following element 108, repr~sents the software to implement the process step of reading in the information or parameters contained in the 5 tables which comprise the database. This information is contained in the file headers of the database, and will be utllized further along in the process. Element 112, which is the element directly following element 110, represents the software to implement the process step of creating the 10 dbEXPRESSTM schema record for each field specified ln the -file headers of the database. This software routlne creates a layout of what is to be done with the specific data contained in the selected tables of the database. The layout provides a method of maintaining a reference between what the 15 format of the data as it was stored in the original database to what it was converted to. The layout is maintained because tlle system user may want to export the dbEXPRESSTM
files into an external file within the database while maintaining the dbEXPRESSTM format. Thus far in the database 20 accessing process, the data contained in the particular ;~
database has been opened so that it can be read into dbEXPRESSTM memory, and a dbEXPRESSTM schema has been created based upon the information contained in the file header.
Element 114, which is the element directly following element 25 112, represents the software to implement the process step of reading a data record from the exposed or raw data into the allocated memory area of dbEXPRESSTM. Element 116, which is the element directly following element 114, represents the software to implement the process step of converting the data 3~ from the database into destination dbEXPRESSTM field record types. No matter the type of data collected, whether it be floating point or alphanumeric, it is converted into true numeric data. For example, if the data in the database is dates of the year represented in ASCII, then the process 2 ~

1 implemented by element 116 converts the ASCII representations of the dates into dbEXPRESSTM representations of the dates.
It is imp,ortant to note that this data is stored or located in program memory space and not computer system memory where 5 the data would be stored as binary data. The data is converted into true numeric data because true numeric data can be manipulated at a much faster rate than any other type or form of data. Element 118, which is the element directly following element 116, represents the software to implement 10 the process step of writing the dbEXPRESSTM records into dbEXPRESSTM data files. This step is a simple reorganizing and compression procedure which allows for the convenient storage of the data. Element 120, which is the element directly following element 118, represents the software to 15 implement the process step of implementing a more records decision loop. If there are more records to process, then this software transfers processing back to element 116 so that the next set of data can be converted into destination dbEXPRESSTM field record types. If there are no more records ~
20 to process, then this software routine is exited and the -software represented by element 122 is then processed.
Element 122 represents the software necessary to terminate the processing of the access database software and return control to the main software routine illustrated in Figure 1.
The next block of software in the overall process which is shown in Figure 1 is element 200 which represents the software to implement the process of building the distribution file or matrix for each table selected in the partlcular database. Element 200 represents a block of !
30 software that comprises a plurality of distribution algorithms which vary for each data type utilized in the particular database selected. For example, two different distribution procedures would be utilized for integer and floating point data types. However, the overall distribution -~
~ ;

, 20~5681 .:.-"

1 building process consists of eight essential steps and can be summarized as follows~
,(1) Determine the number of bins to use for the particular database.
(2) Find the minimum and the maximum values for each field. ~ -(3) Determine the deltabin value.
(4) Fill in the field header values. ;
(5) Assign matrix work areas.

(6) Process the records to build the distribution matrix.

(7) Compress the matrix for this field.

(8) Process all fields in the records contained in the current and relational databases.
15 As stated previously, the primary functions of this distribution is to allow the user to view the data content of the file in a graphical format. The underlying method ~ ~-provides for accessing and displaying the number of records that fall into a given bin and for finding which database -~-20 records are associated with the bin values. A bin is simply a speclfically allocated area of memory.
Figure 3 is the detailed flow chart of the distribution construction routine illustrated in Figure 1 as -~
element 200. Element 202 is the point of entry into the 25 distribution construction routine. Element 204, whlch is the ~
element dlrectly followlng element 202, represents the ~ ~ -software to implement the process of retrieving the names of ;
the particular database and the particular data tables within the particular database. As indlcated previously, the name -3O of the particular database and tables contained therein are retrieved in order to access the data contained ln these -files. However, describing this process as a ~imple retrieval of names is somewhat misleading. The software represented by element 204 is utilized to establlsh a link to 35 the particular database rather than just simply retrieving 2~681 1 names. Element 206, which is the element directly following element 204, represents the software to implement the process step of o~ening the particular database so that the data can be read from the particular database and into dbEXPRESSTM.
5 In addition, element 206 represents ~he software to implement the process step of opening the internal setup or schema of dbEXPRESSTM so that the data from the database may be accurately transferred into dbEXPRESSTM. Element 208, which is the element directly following element 206 and represents 10 the software to implement the process step of retrieving the dbEXPRESSTM schema record created for each field specified in the file header of the database. The schema records are created by the software represented by element 112 in the f low chart of Figure 2. Element 210, whlch is the element 15 directly following element 208, represents the software to :
implement the process step of copylng certain lnformation contalned ln the dbEXPRESSTM schema records directly into the field header for the distribution matrix under construction. :
The information copied includes the field name and the 20 distribution type such as floating point or alphanumeric.
The entire field header structure is discussed in subsequent paragraphs with reference to a figure illustrating the actual layout of the header. Element 212, which is the element directly fo~lowing element 210, represents the software to 25 implement the process step of determining if there are more schema records for this table. If there are more schema ' -~
records for a particular table then this software transfers processing back to element 208 so that the next schema record for the particular table can be retrieved. If there are no .~
30 more schema records for the particular table then this - ::
software transfers processing to the software represented by element 214.
At this point in the processing, the data contained in the records has been readied for further processing. The 35 data and the information from various calculations utilizing , .

2f3~681 ~ ~

1 the data is now utilized to construct the distribution tables for a given table. Element 214, which is the element directly following element 212, represents the software to implement the process step of reading in a record from the -5 dbEXPRESSTM data file. Recall that the dbEXPRESSTM data files were created by writing the dbEXPRESSTM records to the dbEXPRESSTM data files by the software represented by element ~ -118 shown in Figure 2. The reading in of the dbEXPRESS
records is repeated for all records ln a particular field as 10 part of a decision loop. As each record is read, the field values comprising the particular record are evaluated to determine the maximum field value and the minimum field value and which record corresponds to the minimum and maximum field values respectively. This part of the distribution 15 construction process is accomplished be elements 216, 218, and 220. Element 216, which is the element directly following element 214, represents the software to implement the process step of determining the minimum field values as each record is read from the dbEXPRESSTM data file. This 20 software process is a simple procedure which saves a field value as the minimum field value if the value is less than ~ -the previous field value. Element 218, which is the element directly following element 216, represents the software to implement the process step of determining the maximum field 25 value as each record is read from the dbEXPRESSTM data file.
This software process is a simple procedure which saves a field value as the maximum field value if the value is greater than the previous field value. Element 220, which is the element directly following element 218, represents the 30 software to implement the process step of determining if there are more records to process. If there are more records in the dbEXPRESSTM data file then this software transfers processing back to element 214 so that the next record for can be retrieved. If there are no more records then this ' ~: ~:, ' ., ,;- "': "

20~5~81 1 software transfers processing to the software represented by element 222.
,Figure 4 illustrates an example table or sample utilizing eleven records. The field values corresponding to 5 the eleven records were read from the dbEXPRESSTM data files in the order shown in the figure. These numbers are used to illustrate how the distribution tables are to be constructed.
As can be seen by simple inspection of the figure, the minimum field value, MINV, is 7.9, and the maximum field value, MAXV, is 23.2. In actuality MINV and MAXV are determined by the simple process described above, and not by inspection. The actual physical layout of the data is not as shown in Figure 4; however, Figure 4 provides a convenient format for illustrating the distribution construction 15 PrOcess.
Referring back to Figure 3, element 222, which is the element directly following element 220, represents the so~tware to implement the process step of determining the number of bins to use for the database. This software sets 20 the number of bins to 64, 128, 256, 512, or equal to the number of records present. The number 512 is not an upper limit, for example, if increased display resolution is desired the number of bins may be increased to meet user requirements. In this example, the number of bins is set to '-25 11 because there are ll.records. This example represents the simplest implementatlon of dbEXPRESSTM; however, in ;
actuality, the number of blns can vary with the desired level of resolution as stated above. Element 224, which is the element directly following element 222, represents the 30 software to implement the process step of calculating the DELTA~IN value, which is the value to be assigned to each bin, and to assign the bin ranges. The minimum and maxlmum data values and the DELTA~IN value are saved in the field header structure, which is described subsequently. The ~
" '~:'~,' ;' ~. ~ . . . :: .......... .... . . .. . . .. . . . . . .

20~681 : ~-1 DELTABIN value is calculated by the software by utilizing the formula given by .
(MAXV - MINV) / # of bins = DELTABIN. (1) -~
The software has already determined the minimum and maximum ---~
values as well as the setting of the number of bins;
therefore, substituting in the values for MAXV, MINV, and the # of blns results in a DELTABIN equal to ", , DELTABIN = (23.2 - 7.9) / 11 = 1.39. (2) Once the DELTABIN value is calculated by this block of software utilizing equation (1), the bin ranges are 15 determined or assigned by the software by breaking up the -range of values, 7.9 to 23.2, in DELTABIN value increments, ;~
and assigning them to bins in ascending order. The bin ~-~
numbers and thelr respective ranges are shown in Figure 5. -The bin range assignments are a calculated value and are not 20 saved as array elements. Figure 5 is for reference purposes only. The use of these bin ranges will become apparent shortly.
The next step in the distribution construction process is the step of filling in the field header values.
25 As the various information is determined or calculated by the process steps represented in Figure 3, it is eventually stored in the field header data structure. Figure 6 illustrates the basic field header structure. Area 201 is the field name and is filled in by the software process!step ~ -~
30 represented by element 210 of Figure 3. For purposes of this particular example, the field name is chosen to be SAMPLE01.
Area 203 is the distribution type field and contains a two digit code indicative of the particular type of data contained,in the data file, i.e. floating point type data.
35 In this example a distribution type of 5 is utilized and 20~68~

1 shall denote that the field data are floating point values.
Area 203, like area 201 is filled in by the software process step represented by element 210. Area 205 is the number of bins field and contains the actual number of bins utilized 5 for a given field. The number of bins is represented by a hexadecimal number, and in this case since there are eleven bins, hexadecimal B is placed into area 205. The number of bins is set by the software process step represented by element 222 of Figure 3; therefore, it is directly copied 10 into the field header. ~rea 207 is the minimum value field and area 209 is the maximum value field. These two fields ~-contain MINV and MAXV respectively, which were determined by the software process steps represented by elements 216 and 218 respectively. Accordingly, area 207 contains 7.90 and 15 area 209 contains the value 23.20, both of which are ;~
represented by floating point numbers. Area 211 is the delta value field and contains the DELTABIN value calculated by the software process step represented by element 224 of Figure 3;
therefore the number 1.39 is placed there and is represented 20 as a floating point number. Area 213 is the number of bin map words field and contains the number of bin map words utilized. Since the number of bins in this example is set at eleven, only one bin map word will be required. The value of this entry is therefore one. Area 215 ls the bin map area 25 and is utilized to show which particular bins are active. '~
Since there are only eleven bins, a single 16 bit word is needed to show the active bins. This area is filled in as the distribution construction process continues.
At this point in the distribution construction 30 process, the number of bins to use for the database has been determined, the minimum and maximum values of the field have been determined, the DELTABIN value has been calculated and the header values have been filled in to their respective areas. The program will now assign work areas for the 35 distribution ~atrix build. As the matrix is built the work 2o8568~

- 1 9 - ~ ~ :
1 area segments will be expanded on an as needed basis. Upon completion of the distribution construction phase the matrix will be c~mpressed in preparation for the operational phase.
Re~erring back to Figure 3, the next step in the 5 process is assigning the matrix work areas. Element 226, -which is the element directly following element 224, -represents the software to implement the process step of allocating a block of work space for each assigned bin.
Figure 7 graphically illustrates how the memory is divided up 10 for each bin. For each assigned bin the program will allocate a 128 byte block of work space. For this particular example, Figure 7 illustrates the 128 bytes of work space for each of the eleven bins. Table 1 given below contains a listing of the terms or words which comprise the structure of 15 these matrix work areas. The meaning of the words listed in Table 1 shall be explained as the next step in the distribution construction process is explained. : ;~

20 ~
WORD NO. DESCRIPTION MNEMONIC
~ :, .. .
O THE ADDRESS OF THE PREVIOUS MATRIX BIN RECBINFR ;;

BUILD AREA
2 NUMBER OF RECORDS CONTAINED IN THIS BIN RECBINNO ~
3 NUMBER OF RECORD BIT MAP WORDS RECMAPNO ~ -4~lst) STARTING RECORD NUMBER (NEGATED) RECSTRNO ~ ;
3 5(lst) BIT MAP WORDS - (POSITIVE ONLY), THE RECMAPWD
SIGN BIT MUST BE ZERO. THIS BIT MAP ¦ -WORD ALLOWS THE BIT MAPPING OF THE

,, ~ ':
--] The next step in the dlstribution construction process is the processing of the collected records to actually build the distribution matrix. Element 22~ in Figure 3 represents the first block of software in this phase 5 of the process. Element 228, which is the element directly Eollowing element 226, represents the software to implement the process step of reading in a record from the dbEXPRESSTM ;
data file. Accordingly, the software now determines the bin number to update based upon the range for each bin as 10 indlcated in Figure 5. Element 230, which is the element -~
directly following element 228, represents the software to implement the process step of determining the correct bin number. The software represented by this block utilizes the formula given by BIN value = I field value - BIN MIN/DELTABIN 1, (3) -'" ' ' ~ ' ~' wherein the field value and BIN MIN are for the particular record being read at that instant. In addition, the 20 resultant value from equation (3) is rounded up to the nearest whole number. The calculation for the bin number must conclude with the BIN value being set to an absolute value as indicated in the equation. This will take into account the occurrence of negative values. Element 232, 25 which ls the element directly following element 230, represents the software to implement the process step of updating the bin map area in the field header, area 215 ln Figure 6, to indicate that a particular bin is active.
Element 234, which is the element directly following element 30 232, represents the software to implement the process step of updating the information in a particular bin work area. This software is responsible for calculating the values of the words listed in Table 1. Element 236, which is the element directly following element 234, represents the software to 35 implement the process step of allocating an extended work 208~81 1 area iE the current bin work area is larger than 12~ bytes.
Element 238, which is the element directly following element ~-236, represents the software to implement the process of determining if there are more records to process. If there 5 are more records to process, then this soEtware transfers processing back to element 228 so that the next record can be read in from the dbEXP~ESSTM data file. If there are no more records, then this software transfers processing to the software represented by element 240.
To clarify the above process, the procedure is explained with reference to a series of figures which utilize the example values given previously. Figure 8 depicts the processing procedure for record number 1. The record is read from the dbEXPRESSTM data file by the software represented by 15 element 228 in Figure 3. For this particular record, the software process step represented by element 230 utilizes equation (3) to determine the bin number. Substituting in the proper values into equation (3) results in a BIN value given by ;~

BIN Value = l(12.7 - 7.9)/1.39l = 3.45 = 4, (4) since the number is rounded up to the nearest whole number.
Accordingly, the bit corresponding to bin 4 is set in the 25 binmap word in the header. Word 0 in the matrix work area is the address of the previous matrix bin, RECBINFR, and is used if more than 128 bytes of memory is required for a given bln.
In this example, 128 bytes is more than sufficient. Word l in the matrix work area is the address of the extended matrix 30 build area, RECBINT0, and it is only utillzed when any additional memory is required. Accordingly, Word 0 and Word 1 remaln at there lnitial values of 0. Word 2, ls the number of records contained in this bin, RECBINN0. In this example, and at this point in the processing, bin 4 contains 35 one record, record 1, wlth a value of 12.7. Word 3, is the 20~81 ~

1 number of record bit map words, RECMAPNO. This value is initialized to 5 and is updated as more words are utilized.
Word 4, i,s the starting record number, RECSTRNO. This word indicates the specific record number currently being 5 processed and ls represented as a negative number. In this case, slnce the process is on record 1, RECSTRNO is set to -1. Word 5 is the bit map words, RECMAPWD. This word allows the bit mapping of the next 15 records after RECSTRNO. Its use is explained as a description of the next 10 records is 10 9iven.
Figure 9 depicts the processing procedure for -~
record number 2. Once again the record is read from the ;
dbEXPRESSTM data Eile by the software process step represented by element 228 in Figure 3. The software process 15 step represented by element 230 utilizes equation (3) to ~;
determine the correct bin number, which in this case is bin ~
5. ~ccordingly, the bit corresponding to bin 5 is set in the ~E
binmap word in the header. Word 0 and Word 1 remain at there initial values of 0 since more memory space is not needed.
20 Word 2, is the number of records contained in this bin, and since bin 5 contains one record, record 2, with a value of 14.8, Word 2 is set to a value of 1. Word 3, is the number of record bit map words, and remains at the initialized value of 5. Word 4, is the starting record number, and in this 25 case since the process is currently working on record 2, nECSTRNO is set to -2. Word 5 is not used at this point in the processing.
Figure 10 depicts the processing procedure for record number 3. Once again the record is read from the 30 dbEXPRESSTM data file by the software process step represented by element 228 in Figure 3. The software process step represented by element 230 utilizes equation (3) to determine the correct bin number, which in this case is bin 8. Accordingly, the bit corresponding to bin 8 is set in the 35 binmap word in the header. Word 0 and Word 1 remain at there 20~681 1 initial values of 0 since more memory space is not needed.
Word 2, is the number o~ records contained in this bin, and since bin,8 contains one record, record 3, with a value of 17.9, Word 2 is set to a value of 1. Word 3, is the number 5 oE record bit ~ap words, and remains at the initialized value of 5. Word 4, is the starting record number, and in this ~
case since the process is currently working on record 3, ~ -RECSTRNO is set to -3. Word 5 is not used at this point in the processing. - -~
Figure 11 depicts the processing procedure for record number 4. Once again the record is read from the dbEXPRESSTM data file by the software process step represented by element 228 in Figure 3. The software process step represented by element 230 utilizes equation (3) to 15 determine the correct bin number, which in this case is bin 11. Accordingly, the bit corresponding to bin 11 is set in the binmap word in the header. Word 0 and Word 1 remain at there initial values of 0 since more memory space is not needed. Word 2, is the number of records contained in this 20 bln, and since bin 11 contains one record, record 4, with a value of 22.4, Word 2 ls set to a value of 1. Word 3, is the number of record bit map words, and remains at the initialized value of 5. Word 4, is the starting record number, and in this case since the process is currently 25 working on record 4, RECSTRN0 is set to -4. Word 5 is not used at this point in the processinq.
Figure 12 depicts the processing procedure for record number 5. Once again the record is read from the dbEXPRESSTM data file by the software process step 30 represented by element 228 in Figure 3. The software process step represented by element 230 utilizes equation (3) to determine the correct bln number, which in this case is bin 11 again. Since the bit corresponding to bin 11 is already set in the binmap word in the header, there is no need to set ~-35 it again. Word 0 and Word 1 remain at there initial values .,-. ~ .' 20~5681 1 of 0 since more memory space is not needed. Word 2, is the number of records contained in this bin, and since this is the secon~ record in bin 11, record 5, with a value of 22.4, Word 2 is set to a value of 2. Word 3, is the number of 5 record bit map words, and is set to a value of 6 because Word 5 is now going to be utilized. Word 4, is the starting record number for this particular bin and it remains the same as it was for processing record 4 because the first record in ~-5-~
bin 11 is still record 4. Word 5 is set to a value of 0001.
lO As was stated previously, this number allows the bit mapping ~ ~-of the next 15 records after ~ECSTRNO. Therefore, the value 0001 is utilized to indicate record 5 because 0001 is the i~
first position after record 4 which is indicated by the value ~i of RECSTRNO. The previous values for Words 0 through 5 are 15 shown in the figure; however, only the second column would be saved if there were no more records.
Figure 13 depicts the processing procedure for xecord number 6. Once again the record is read from the dbEXPRESSTM data file by the software process step 20 represented by element 228 in Figure 3. The software process step represented by element 230 utilizes equation (3) to determine the correct bin number, which in this case is bin 4 ;
again. Since the bit corresponding to bin 4 is already set in the binmap word in the header, there is no need to set it ;~
25 again. Word 0 and Word 1 remain at there initial values of 0 -since more memory space is not needed. Word 2, is the number of records contained in this bin, and since this is the second record in bin 4, record 6, with a value of 12.7, Word 2 is set to a value of 2. Word 3, is the number of record 3o bit map words, and is set to a value of 6 because Word 5 is now going to be utilized. Word 4, is the starting record number for this particular bin and it remains the same as it was for processing record 1 because the first record in bin 4 ;~ -~
is still record 1. Word 5 is set to a value of 0010. The ~-35 value 0010 is utilized to indicate record 6 because 0010 is ',':"~'" '~ ,' . ~ ' :, .., ..:,,: ;:
:~' ~ " ''~ ' ' : ': , , 2D~5~81 ~

l the fifth position after record 1 whlch is indicated by the value of RECSTRNO. Once again, the previous values for Words 0 through,5 are shown in the figure; however, only the second column would be saved if there were no more records.
Figure 14 depicts the processing procedure for record number 7. Once again the record is read from the dbEXPRESSTM data file by the software process step represented by element 228 in Figure 3. The software process step represented by element 230 utilizes equation (3) to determlne the correct bin number, which in this case is bin 4 again. Since the bit corresponding to bin 4 is already set ln the binmap word in the header, there is no need to set it again. Word 0 and Word l remain at there initial values of 0 since more memory space is not needed. Word 2, i5 the number 15 ~f records contained in this bin, and since this is the third record in bin 4, record 7, with a value of 12.7, Word 2 is set to a value of 3. Word 3, is the number of record bit map words, and is set to a value of 6 because Word 5 is now going to be utilized. Word 4, is the starting record number for 20 this particular bin and it remains the same as it was for processing record 1 because the first record in bin 4 is still record 1. Word 5 is set to a value of 0030. The value 0030 is utilized to indicate record 7 because 0030 is the sixth position after record 1 which is lndicated by the value 25 of RECSTRNO. Once again, the previous values for Words 0 through 5 are shown in the figure; however, only the third column would be saved if there were no more records.
Figure 15 depicts the processing procedure for record number 8. Once again the record is read from the 30 dbEXPRESSTM data file by the software process step represented by element 228 in Figure 3. The software process step represented by element 230 utilizes equation ~3) to determine the correct bin number, which in this case is bin 5 again. Since the bit corresponding to bin 5 is already set 35 in the binmap word in the header, there is no need to set it 20~568~

1 again. Word 0 and Word 1 remain at there initial values of 0 since more memory space is not needed. Word 2, is the number of records contained in this bin, and slnce thls is the second record in bin 5, record 8, with a value of 14.8, Word ;
5 2 is set to a value of 2. Word 3, is the number of record bit map words, and is set to a value of 6 because Word 5 i9 now going to be utilized. Word 4, is the starting record number for this particular bin and it remains the same as it was for processing record 2 because the first record in bin 5 lO is still record 2. Word 5 is set to a value of 0020. The value 0020 is utilized to indicate record 8 because 0020 is the sixth position after record 2 which is indicated by the value of RECSTRNO. Once again, the previous values for Words 0 through 5 are shown in the figure; however, only the second 15 column would be saved if there were no more records.
Figure 16 depicts the processing procedure for -~
record number 9. Once again the record is read from the dbEXPRESSTM data file by the software process step represented by element 228 in Figure 3. The software process 20 step represented by element 230 utilizes equation (3) to determine the correct bin number, which in this case is bln 9. Accordingly, the bit corresponding to bin 9 i9 set in the binmap word in the header. Word 0 and Word 1 remain at there initial values of 0 since more memory space is not needed.
25 Word 2, i9 the number of records contained in this bin, and since bin 9 contains one record, record 9, with a value of 19.7, Word 2 is set to a value of 1. Word 3, is the number ;~
of record bit map words, and remains at the initialized value of 5. Word 4, is the starting record number, and in this 3O case since the process is currently working on record 9, RECSTRNO is set to -9. Word 5 is not used at this point in the processing.
Figure 17 depicts the processing procedure for record number 10. Once again the record is read from the 35 dbEXPRESS data file by the software process step 2 0 8 ~ 6 ~

l represented by element 228 in Figure 3. The software process step represented by element 230 utilizes equation (3) to determine~ the correct bin number, which in this case is bin 1. Accordingly, the bit corresponding to bin 1 is set in the 5 binmap word in the header. Word 0 and Word 1 remain at there initial values of 0 since more memory space is not needed.
Word 2, is the number of records contained in this bin, and since bin 1 contains one record, record 10, with a value of 7.9, Word 2 is set to a value of 1. Word 3, is the number of -l0 record bit map words, and remains at the initialized value of 5. Word 4, is the starting record number, and in this case since the process is currently working on record 10, RECSTRNO
is set to -10. Word 5 is not used at this point in the processing.
Figure 18 depicts the processing procedure for record number 11. Once again the record is read from the dbEXPRESSTM data file by the software process step represented by element 228 in Figure 3. The software process step represented by element 230 utilizes equation (3) to 20 determine the correct bin number, which in this case is bin 1 again. Since the bit corresponding to bin 1 is already set in the binmap word in the header, there is no need to set it again. Word 0 and Word 1 remain at there initial values of 0 since more memory space is not needed. Word 2, is the number 25 of records contained in this bin, and since this is the second record in bin 1, record 11, with a value of 8.6, Word -~
2 is set to a value of 2. Word 3, is the number of record bit map words, and is set to a value of 6 because Word 5 is now going to be utilized. Word 4, is the starting record 30 number for this particular bin and it remains the same as it was for processing record 10 because the first record in bin 1 is still record 10. Word 5 is set to a value of 0001. The value 0001 is utilized to indicate record 11 because 0001 is -~
the first position after record 10 which is indicated by the 35 value of RECSTRNO. Once again, the previous values for Words 2 0 ~ ~ ~ 8 ~

1 ~ through 5 are shown in the figure; however, only the second column would be saved if there were no more records.
,The next step in the distribution building process is the compression of the field header information and the 5 information for every field. As stated previously, the original allocation of memory was in 128 byte blocks. The purpose of this step is to eliminate unused space from the distribution matrix. Referring back to Figure 3 once again, element 240, which is the element directly following element 10 238, represents the software to implemented the process step of compressing the data. The software basically just eliminates the unused spaces of memory. F3gure 19 represents the final data content of the bin work area as created from the final values, i.e. last column of values, from Figures 8 15 through 18. The compression process removes the record link words, REC~INFR and RECBINTO, by ad~ustlng RECMAPNO, which is the number of words in the bin matrix. The data is then compressed into a contiguous block of memory as shown in Figure 20. Element 242, which is the element directly 20 followlng element 240, represents the software to implement the process step of writing the compressed distribution ~ ' records to the distribution file where it can be utilized for display purposes as illustrated in the bar chart of Figure 23, which is explained in subsequent paragraphs. The entire 25 display procedure is also explained in subsequent sections.
Element 244, which is the element directly following element 242, represents the software to implement the process step of deallocating the memory utilized to build the matrix distributions. The flnal step in the distribution building 30 process is the processing of all fields in the records contained in the current and relational databases. This means that the entire process is repeated until all fields are processed. Element 246, which is the element directly following element 244, represents the software to implement ~;
35 the process step of terminating the distribution building '.'''''' ~'"'~

2 ~ 8 ~

l process and return processing control to element 300 in Figure 1. -~ The next block of software in the overall process which is shown in Figure 1 is element 300 which represents 5 the software to implement the process of graphically displaying the selected fields for which the distribution matrix was built. This software is for the primary display of the records for selected fields. The graphical dlsplay is accomplished utilizing a variety of graphical display formats lO including bar charts, line charts, scatter representations, pie charts and nyquest curves. The software includes two basic steps. The first step in the process is opening a specific dbEXPRESSTM file for display and is represented by element 302 in Figure 1. The second step in the process is 15 expanding the compressed data contained in the particular dbEXPRESSTM data file so that it can be displayed, and is represented by element 316 in Figure 1. The software process represented by element 302 is further described with reference to the flow chart of Figure 21 and the software 20 process represented by element 316 is further described with reference to the flow chart of Figure 22. In addition, figures illustrating the types of primary displays are also described.
Referring now to Figure 21, there is shown a 25 detailed flow chart representing the software process to implement the dbEXPRESSTM file opening software. Element 304 i9 the point of entry into the dbEXPRESSTM file opening routine. Element 306, which is the element directly following element 304, répresents the software to implement 30 the process step of opening the specified distribution and data files so that the particular files can be accessed.
Element 308, which is the element directly following element 306, represents the software to implement the process step of allocating the distribution file header record structure.
35 This software simply reserves a block of memory for the 2~6~

l requ~red information to be displayed because of the unknown size or quantity of the data contained therein. Element 310, which is ~he element directly following element 308, represents the software to implement the process step of 5 reading the information stored in the distribution file header record. Recall that the file header contains such information as the field name, the distribution type and the number of bins utilized. Element 312, which is the element dlrectly following element 310, represents the software to lO implement the process step of returning the distribution handle. This software is responsible for moving the pointer to a ready position so that the next display can be accomplished. The linkage for data table access is the ~ ~
distribution handle or pointer. Element 314, which is the ' .'~;
15 element directly following element 312, represents the software to implement the process step of terminating processing of the opening routine and passe~ processing control back to the display software represented by element 300 in Figure 1. ~-Referring now to Figure 22 there i9 shown a detailed flow chart representing the software process to implemént the expansion of compressed data software. Element 318 is the point of entry in the expansion software. Element -320, which is the element directly following element 318, 25 represents the software to implement the process step of reading the data contained in the compressed distribution record for the specified or selected field. The memory space was previously allocated by the software represented by element 308 in Figure 21. Element 322, which is the element 30 directly following element 320, represents the software to implement the process step of building and expanding a ;~
distribution record from the compressed record so that it can be properly displayed by the graphics software process ' represented by element 300 in Figure l. Element 324, which 35 is the element directly following element 322, represents the -2~5S8~

1 software to implement the process step of terminatlng processing of the expansion routine and passes control back to the di$play software represented by element 300 of Figure 1.
Figure 23 illustrates a graphical presentation of fictitious customer records, utilizing a bar chart display format, which provides an example of the graphical display of a distribution available to the user of this method. The bar charts in this figure represent a primary display of the 10 distribution for all selected fields of customer records. In a primary display, the color of the background is grey and the bar chart showing the number of records is in red. The abscissa represents the first letter of the various customer names, and the ordlnate represents the number of customers or 15 customer records. Accordingly, the first bar in the chart indicates that there are 70 customer names starting with the letters A, B, or C in the database, the second bar indicates that there are 30 customer names starting with the letters D
through F, the third bar indicates that there are 10 customer -~
20 names starting with the letters G through I,the fourth bar indicates that there are 40 customer names starting with the letters J through L, the fifth bar indicates that there are 20 customer na~es starting with the letters P through R, the sixth bar indicates that there are 100 customer names 25 starting wlth the letters S through U, the seventh bar indicates that there are 10 customer names starting with the letters V through X, and the eighth bar indicates that there are 20 customer names starting witn the letters Y and Z.
Figure 23 illustrates a primary graphical display 30 of fictitious customer records and provides a basis for further description of the detailed selection and display process. Before the detailed selection and display process illustrated in Figures 26, and 29 through 32 is described, the functional relationship between a primary graphical 35 display and the distribution matrix from which it was 2~85~1 1 generated will be described with respect to Figures 20 and 24.
~ Figure 24 illustrates a primary graphical display generated from the compressed distribution matrix represented 5 in Figure 20. In bin 1, RECBINNO, which is the number of records contained in the particular bin is equal to 2, ~ - ~
accordingly the first bar in Figure 24 shows that there are ~ ~-two records. Recalling Figure 5, bin 1 contains or represents record data ranging in value from 7.90 to 9.29;
10 therefore, the ordinate in the display of Figure 24 shows two records, whatever they may represent, having values which are ;
in a range of between 7.90 and 9.29 in the particular database which has been accessed. In bin 4, as shown in Figure 20, ~EC~INNO is equal to 3, accordingly, the second 15 bar in Figure 24-shows that there are three records within the specified range. The range of record data in this particular bin, as shown in Figure ~, is between 12.09 and ' 13.48. In bin 5, RECBINNO is equal to 2, accordingly, the third, bar in Figure 24 shows that there are two records within the ~pecified range. The range of record data in this particular bin, as shown in Figure 5, is between 13.49 and 14.88. In bin 8, RECBINNO is equal to 1, accordingly, the fourth bar in Figure 24 shows that there is one record within the specified range. The range of record data in thls 25 particular bin 1 as shown in Figure 5, is between 17.69 and 19.08. In bin 9, RECBINNO is equal to 1, accordlngly, the fifth bar in Figure 24 shows that there is one record within -- i~
the specified range. The range of record data in this particular bin is 19.09 to 20.48. In bin 11, RECBINNO ls 30 equal to 2, accordingly, the sixth and final bar in Figure 24 shows that there are two records within the specified range.
The range of data in this particular bin, as shown in Figure 5, is between 21.89 and 23.28. Each of the programs or software routines used to generate the graphical displays 35 hereinafter illustrated and described with respect to Figures ..';~, 2~5~1 1 25 through 32 utillze compressed distribution matrices to obtain data for the display in a manner similar to the way the progr~m generated the graphical display in Figure 24 from the compressed matrix illustrated in Figure 20.
The actual data contained in the selected database ls not necessary ~or a primary graphical display. The prlmary graphical display as well as the secondary graphical display provides a useful and informative method of viewing and/or studying the overall distribution of the data without 10 being burdened with the actual data. As can be seen from Figure 24, only the number of data records over specified ranges are illustrated, not the actual values. If and when actual data is needed, the system user simply requests a detail display, which is described and illustrated in 15 subsequent paragraphs, and the actual data represented by the data records in the primary graphical display is displayed in a list box. The actual data is stored in system memory or extended system memory and is easily accessed as explained in the description of the detailed graphical display procedure.
Returning to the detailed description of Figure 23, the user may now select any bar for detailed selection. A
detailed selection represents a secondary display of the data. Element 400, which is the element directly following element 300 in Figure l, represents the software to implement 25 the selection process. The software provides the record number or numbers for the user selected bins. Element 500, which is the element directly following element 400, represents the software to implement the process step of displaying the secondary display. Figure 25 is a detailed 30 flow chart representing the software process to perform the secondary display. Element 502 is the point of entry into the secondary display routine. Element 504, which is the element directly following element 502, represents the software to implement the process step of readin~ the 35 compressed distribution record for the specified primary 2~5681 1 field. This software reads the compressed distribution for all records in the field that was previously displayed as -illustra~ed in Figure 23. Element 506, which is the element directly following element 504, represents the software to 5 implement the process step of reading the compressed distribution record for the specified secondary field.
Basically, this software process is responsible for locatlng the distribution just for the specified records in that field. The records were chosen by the user and implemented 10 by the software process represented by elements 400 and 500 in Figure l. Element 508, which is the element directly following element 506, represents the software to implement the process step of building and expanding the secondary distribution record from the compressed record so that it can 15 be properly displayed by the graphics software process represented by elements 400 and 500 in Figure 1. Element 510, which is the element directly following element 508, represents the software to implement the process step of - . . .
terminating processing of the secondary display routine.
Referring now to Figure 26, there is shown a secondary display of the first bar of the bar chart in Figure 23. The secondary display, as stated previously, is a result of the selection made by the user. The background color of the display ls still gray, and the main bar graph or chart is 25 still red; however, the selected bar, the one indicated by X's instead of REC's, i8 now a blue color to indicate that a selection has been made. In addition, a new chart representing the details of the selected bar is placed along the bottom of the display screen in blue. The detalled chart 30 located at the bottom of the display breaks down the original ~
bar, selected by the user, into a detailed graphical display. ~ -For example, the detailed chart illustrates the four customer companles and how many records for each of the named companies, whereas the original bar chart only illustrates 2~85681 : ::

1 that there were a total of 70 records for companies starting with the letters A, B, and~or C.
,The user may now select a single bar or bars from the detail graph. A selection from the detail graph i8 5 referred to as a tertiary display. In a tertiary display, the user may also request that the selected data be presented as a list box with further selection capability. Referring -once again to Figure l, element 600, which is the element directly following element 500, represents the software 10 process to display the selected records as a tertiary display and also to provide a list box if required. Figures 27 and 28 are detailed flow charts representing the software processes to display the selected tertiary display and provide a list box for the selection. Figures 27 and 28 15 represent identical code or software; however, the code represented by the flow chart in Figure 27 is for the secondary records while the code represented by the flow chart in Flgure 28 is for the tertiary records. The elements of the flow chart in Figure 28 are numbered with identical 20 numbers as those in the flow chart of Figure ~7 except with the addltion of primes. Any differences in the code are indicated. Element 602 is the point of entry lnto the tertiary display routine. Element 604, which is the element directly following element 602, represents the software to ~-25 implement the process step of reading in the compre~sed distribution records utilized for the secondary display for the specified field. In Figure 28, element 604' represents the software to implement the process step of reading in the compressed distribution records utilized for the tertiary 30 display for the specified fleld. Element 606, which is the element directly following element 604, represents the software to implement the process step of determining the specific record number and field values for the specified bins so that the data may be displayed. Element 60B, which 35 is the element directly following element 606, represents the - 20~5~81 ~

1 software to implement the process step of sorting the record ~ -numbers and field values for use in creating a 11st box of the selec~tions made. This software is a basic sort routine -~
or utility which sorts the record numbers and field values in 5 ascending field value order. The use of the field listing is made apparent in the description of a tertiary display.
Element 610, which is the element directly following element 608, represents the software to implement the process step of ;
terminating the processing of this routine and passes 10 processing control to the software represented by the flow chart of Figure 28.
Figure 29 provides an example of a tertiary ~--display. The background color is still grey, the main bar chart is still red; however, blue is utilized to display the 15 selected records from the detailed bar. In this case, the - ~;~
first 30 entries in the first bar have been selected, accordingly, they appear in blue on the display and are represented as X's in the bar as opposed to REC's. In addition, the selected companies, ARTLER and BUTLER, have 20 their records in the detailed chart appear in blue, as indicated by X's, while the other two companies appear in red ~ -as indicated by the REC's. Also, the listing box which provides a listing of the companies and the number of records associated wlth each is provided in two colors; namely blue -~
25 for the selected companies and red for the others. The only limiting factor on what level display can be made is the number of records in a bin. Basically, what this means is that a detailed display can be made up to the last record in a bln or field.
Figure 30 illustrates another aspect of the distribution indexing methodology. Figure 30 illustrates the number of records that contain a sale made by a given salesman. The highlighted or blue records indicated by the X's in the figure, are the identification of which sales were 35 made to the selected companies shown in Figure 26, i.e.

2~1~5681 1 companies having their names beginnlng with the letters A, B, and/or C. Basically, this figure illustrates the effect of the sele~tion process made in Figure 26. If no selection had been made to get to Figure 26,then all the salesman records 5 would appear in red, REC's, or not highlighted. Therefore, as illustrated by this figure, the selection of particular companies in one display shows up on the display of different items such as in thls case, salesmen in a different display.
The system user further has the capability of making a 10 selection to eliminate a given salesman. Figure 31 illustrates the display after salesman 1, S1, has been eliminated. Accordingly, the number of total sales has been reduced by 2, which is the number of sales salesman 1 was xesponsible for as indicated in Figure 31. The effect of the 15 selection to eliminate salesman 1 or to deselect him can be seen in the customer records display as shown in Figure 32.
As can be seen in this figure, the customer records have been reduced to account for the loss of salesman 1. Therefore, one can see that the selection process works in both 20 directions. The effect of a selection or deselection alters the related distributions, thereby effecting the results in all other related displays.
The use of this distribution indexing methodology is extremely powerful in that the effect of various 25 selections and deselections are reflected in all other data fields which comprise a given database. The methodology also comprises a "what if" and "commit" mode of operation. For example, if instead of actually making a specific selection, the system user would only like to see what the effect of the 30 selection would be without actually making the change, he or she could make a "what if" selection. If the "what if"
selection is made, the "what if" selected data is displayed in yellow, and the user can determine if this is what he or she desires to do. If it is what the user wants, then the 35 user hits the "commit" command and the selection is made and , .

2 ~ 8 5 B ~

1 displayed in blue as before. If this is not what the user wants, then a different ~what if" can be done or a different ;
selection,can be made.
Referring now to Figure 33, there is shown a high 5 level flow chart representing the dbEXPRESSTM graphical interface procedure described in detail in the preceding -paragraphs. The software for monitoring the dbEXPRESSTM ~ ~
graphical interface is primarily event driven. When an ~ -event occurs, such as selecting a button, dbEXPRESSTM reacts 10 to the event. The action that is performed by dbEXPRESSTM
depends upon which mode of operation dbEXPRESSTM is currently processing. The dbEXPRESSTM software is capable of operating in one of three modes. Mode 1 of operation is the primary graph display mode, mode 2 of operation is the detail graph display mode, and mode 3 of operation is the list box display ;
mode.
In dbEXPRESSTM mode 1, which is implemented by the software represented by element 300 in Figure 1, the system user can view any field dlstribution graph by selecting a 20 field from a selection list box and selecting the display button. When this event occurs, the software will implement ;~
a process whlch will remove the current field distribution graph, if any is already displayed, and replace it with the selected fields primary distribution graph. Elements 802, 25 804, 806, and 808 of Figure 33 represent the software ;~
process' to implement the mode 1 procedure. Element 802 represents the software process to allow the system user to select a field or fields from the selection list box so that -they may be displayed. Element 804, which is the element 30 directly following element 802, represents the software to implement the process step of opening the selected distribution table(s) and retrieve the desired distribution so that it may be displayed. Element 806, which is the element directly following element 804, represents the 35 software necessary to display the primary graph of the - ' 208568~ .

1 selected fields. Element 808, whlch is the element directly following element 806, represents the software to lmplement the pro~ess step of displaying the fields not selected in the selection list box. The remaining fields displayed ln the 5 selection list box are displayed so that the system user can make additional selections for primary graphical display.
Each selection for primary graphical display results in a new graph as indicated above. Elements 802, 80~, 806, and 808 provide a general description of the process lmplemented by lO element 300 in Figure 1 and its associated detalled flow charts of Figures 21 and 22.
The remaining elements in Figure 33 represent the software necessary to implement the processing for all three modes of operation. Element 810, which is the element 15 directly following element 808, represents the software to implement the process step of initializing the mode count.
Element 812, which is the element directly following element 810, represents the software to implement the process step of monitoring the graphical user interface to determine which 20 mode of operation to enter into. Element 814, which is the element dlrectly following element 812, represents the software to implement the processing of the events which occur in mode 1 of dbEXPRESSTM operation. The events for mode 1 of operation have been described above. Element 816, 25 whlch is the element directly following element 814, represents the software to implement the processing of the ;
events which occur in mode 2 of dbEXPRESSTM operation.
Element 818, which is the element directly following element 816, represents the software to implement the processing of 3O the events which occur in mode 3 of dbEXPRESSTM operation. ~- Details of event processing for modes 2 and 3 of dbEXPRESSTM
operation are given in the following paragraphs and are related to the various previously described detailed flow charts.

-- 208~681 -40- ~-1 In mode 2 of operation, the system user has the ability to expand the data display for a given bar, lf a bar graph is ujtilized as the display, ln the primary display and to select and deselect all of the records for any bar being 5 displayed. Thls processing is implemented by the software ;~
processes represented by elements 400 and 500 in Figure 1.
The processing for selecting and deselecting a bar is done by moving the mouse to a bar and clicking or depressing the right button of the mouse. It should be noted, however, that 10 the mouse is utilized only to illustrate the procedure. In actuality, the selection itself can be made through a ~ -keyboard operation, a touch screen button, or even a light pen operation. Regardless of how the actual selection is made, the software will retrieve the bin selection count.
15 If the count is greater than 0, then the selection count is set to 0 and the particular bar is shaded in red, indicating that no selection has been made. If on the other hand, the count is equal to 0, then the selection count is set to the ~
total count and the bar is shaded in blue indicating that a ~-20 selection has been made. The software process will then update, for every field, the bin records selection count.
The software responsible for this processing is represented ~ ~
by the detailed flow chart in Figure 25. The software ~ -process represented by the flow chart in Figure 25 i~ called 25 for every field indicating the primary fields select bin records count and the total number of selected records.
The technique for the expansion of the data as described above is accomplished as follows. For every primary bar graph, there exists for each bar in the graph a 30 range of bin numbers, the coordinates of the bar, which is the bar hot region for placement of the mouse, the number of selected records, "what if" records, and the total number of records in or represented by the bar. When the system user selects a bar, an event will occur. The bar is selected by 35 posltioning the mouse in a bar hot spot region and clicking -,~' ' "~
'.

2 ~ 8 1 1 or depressing the left button. When the left button i9 clicked, the dbEXPRESSTM mode will change from mode 1 to mode 2 and the~primary display graph will be removed. The mode 2 software will retrieve the selected bar bin range, build the 5 detail graph and then display the graph. In mode 2, the software will monitor for three specific mode events. This software process step is represented by element 816 in Figure 32. Event 1 of mode 2 i5 simply a return to the primary display graph. This event occurs when the mouse is 10 positioned anywhere in the graph region and the left button on the mouse is clic~ed or depressed. The software will change the present mode of operation back to mode l, remove the currently displayed graph and redraw the primary graph.
Event 2 of mode 2 is the select/deselect optlon for a given 15 or selected bar in a graph. This event occurs when the right button of the mouse is clicked on and the mouses cursor position is in a particular bar's hot region. The actual process for selecting and deselecting has been previously descrlbed in the proceeding paragraphs. Event 3 of mode 2 is 20 a display selected bar's records option. This event occurs when the system user selects the select display bar value button and the particular bar. This is done by clicking the mouse on the display button, positioning the mouse on the partlcular bar's hot region and depressing the right button.
25 When this event occurs, the software process will remove the detail display box, and set the mode to mode 3.
When mode 3 is initiated, the software process represented by the flow chart in Figure 27 will retrieve the records for the selected bar and display the records in!a 30 list box. When the retrieval software, represented by the flow chart on Figure 27, is called to retrieve the data, the information required is the bin range of the bar, the field distribution record, and the select record data retrieve mode ~
~selected records or not-selected records). The retrieved ~-35 software wll1 rntrleve the record lndex and value~ contalned 2 ~ ~ 5 6 ~

l within the bin range for the selected bar. The values of the ~ ~
or corresponding to the record will then be displayed in the ~ -list box.~ It i9 important to note that up to this point in the processing the only input/output operation that has 5 occurred was when the distribution table was generated and the distribution records for each field was read. -~
There are two events that are monitored while in mode 3. The software allows the system user to select or deselect any records in the list box simply by positioning lO the mouse cursor on a record and depressing the left button. ~
When the record is selected or deselected it will be shaded ~;
in either blue, which indicates a selection has been made, or . . . ~. . .
red, which indicates that no selection has been made. The system user may then perform two events; namely, select the 15 cancel button or select the commit button. When the cancel event occurs the list box and buttons will be removed, the mode will be set to mode 2, and the detail graph from mode 2 wlll be redrawn. The commlt event will lnvestigate that list box setting and update the selected bars bin count. After 20 the select code for the bar has been updated the list box will be removed and the detail graph will be redrawn and the mode set to mode 2. The event processing is accomplished by the software process step represented by element 818.
Referring back to Figure l once again, the ~
25 remaining elements are now described. Element 700, which i8 :
the element directly following element 600, represent~ the software process to provide the users with various user selection output modes for the graphical displays. The output modes for the displays can be in tabular format 30 output, two-dimen~ional graphic output, and three-dimensional ;
graphic output. The software also provides for the outputting of the data. The data can be output to a disk 702, a printer 704, a CRT 706 or a local area network or a modem 708. The user makes the partlcular selections. ;~
~
,~ :,' .

~ ~ , , 1 .. ,, ,''~.~." ,'~ ', "~ "~ - ''"~ " '~ " ~ ; """~ "~ "~

,, . . ,,,, A . .,. ~.j - ,, ,. , ' i, ; . .

1 While the invention has been particularly shown and descrlbed wlth respect to the preferred embodiments thereof, it should,be understood by those skilled in the art that the foregoing and other changes in form and details may be made 5 therein without departing from the spirlt and scope of the invention which ~hould be limited only by the scope of the appended claims.

~:

''',' . .:' ~
. ' ,. '.' ~ ' ; ',' "., ;

Claims (14)

THE EMBODIMENTS OF THE INVENTION IN WHICH AN EXCLUSIVE
PROPERTY OR PRIVILEGE IS CLAIMED ARE DEFINED AS FOLLOWS:

1. A method for selecting and displaying a distribution of data in a selected data structure, said selected data structure having a particular file structure associated therewith and comprising at least a plurality of data records, and a plurality of data fields having data therein, said method comprising the steps of:
(a) accessing said selected data structure resident on a host system and reading the data therein to determine selected parameters of the distribution of said data;
(b) building a plurality of distribution matricies from data contained within said selected data structure with each matrix having a field header and an array of data bins and data summeries associated therewith (c) generating a display from said plurality of distribution matricies for said selected data structure to provide a graphical representation of the distribution of data contained within said selected data structure.

2. The method for selecting and displaying according to Claim 1, wherein said step of accessing said selected data structure comprises the steps of:
(a) retrieving each of said plurality of data records within said selected data structure;
(b) reading each data field in one of said plurality of data records, and converting the data in said data field to a distribution data file; and (c) repeating step (b) for each of said plurality, of data records to form a master distribution data file for said selected data structure.

3. The method for selecting and displaying according to Claim 2, wherein said step of building a plurality of distribution matricies comprises the steps of:
(a) calculating the number of bins to utilize for all data within said selected data structure;
(b) calculating a minimum and a maximum field data value for said plurality of data records in said selected data structure;
(c) calculating a deltabin value for one of said plurality of data fields in said selected data structure and assigning a bin range for the determined number of bins based upon said deltabin value;
(d) generating a field header for one of said plurality or data fields from said distribution data file and at least the results of steps (a)-(c);
(e) processing all data contained in said master distribution data file to build said plurality of distribution matrices from said field header and said master distribution data file;
(f) processing each of said plurality of data fields contained in said selected data structure by repeating steps (c) through (e).

4. The method for selecting and displaying according to Claim 3, wherein said step of building a plurality of distribution matricies further comprises the step of compressing all information contained in said plurality of distribution matricies into a master distribution matrix.

5. The method of selecting and displaying according to Claim 4, wherein said step of generating a display from said plurality of distribution matricies for said selected data structure comprises the steps of:
(a) opening at least said master distribution matrix to be displayed; and (b) displaying the distribution of data contained in said master distribution matrix in a graphical format.

6. The method of selecting and displaying according to Claim 5, wherein said step of generating a display from said plurality of distribution matricies for said selected data structure further comprises the step of expanding the compressed information contained in said master distribution matrix.

7. A method for selecting and displaying a distribution of data in a selected data structure, said selected data structure having a particular file structure associated therewith and comprising at least a plurality of data records, and a plurality of data fields having data therein, said method comprising the steps of:
(a) accessing said selected data structure resident on a host system and reading the data therein to determine selected parameters of the distribution of said data;
(b) building a plurality of distribution matricies from data contained within said selected data structure with each matrix having a field header and an array of data bins and data summeries associated therewith to form a master distribution matrix;
(c) generating a first display from said master distribution matrix for said selected data structure to provide a graphical representation of the distribution of data contained within said selected data structure, said first display providing a range of data values for each selected field; and (d) modifying a specific range of data values for a selected data field in said first display, and thereafter generating a second display to determine the representative effect of said modification upon the master distribution matrix.

8. The method of selecting and displaying according to Claim 7, wherein said method further comprises the step of generating a third display of a detailed distribution for a selected range of said master distribution matrix to provide a detailed graphical representation of the data contained within the selected range of said master distribution matrix.

9. The method for selecting and displaying according to Claim 8, wherein said step of accessing said selected data structure comprises the steps of:
(a) retrieving each of said plurality of data records within said selected data structure;
(b) reading each data field in one of said plurality of data records, and converting the data in said data field to a distribution data file; and (c) repeating step (b) for each of said plurality of data records to form a master distribution data file for said selected data structure.

10. The method for selecting and displaying according to Claim 9, wherein said step of building a plurality of distribution matricies comprises the steps of:
(a) calculating the number of bins to utilize for all data within said selected data structure;
(b) calculating a minimum and a maximum field data value for said plurality of data records in said selected data structure;
(c) calculating a deltabin value for one of said plurality of data fields in said selected data structure and assigning a bin range for the determined number of bins based upon said deltabin value;

(d) generating a field header for one of said plurality of data fields from said master distribution data file and at least the results of steps (a)-(c);
(e) processing all data contained in said master distribution data file to build said master distribution matrix from said field header and said master distribution data file;
(f) processing each of said plurality of data fields contained in said selected data structure by repeating steps (c) through (e).

11. The method of selecting and displaying according to Claim 10, wherein said step of generating a first display from said master distribution matrix for said selected data structure comprises the steps of:
(a) opening at least said master distribution matrix to be displayed, and (b) displaying at least a portion of the distribution of data contained in said constructed distribution matrix in a graphical format.

12. The method for selecting and displaying according to Claim 7, wherein said step of modifying a specific range for a selected data field in said first display further comprises generating a second master distribution matrix wherein said master distribution matrix is altered by said modification of said specific range of data values to form said second master distribution matrix.

13. The method for selecting and displaying according to Claim 9, wherein the step of generating a third display further comprises the steps of:
(a) identifying the data records in the master distribution data file having field values within said modified range;

(b) partitioning a listbox within said third display; and (c) displaying the field values of the records identified in step (a) within said listbox.

14. The method for selecting and displaying according to Claim 12, wherein the step of generating a third display further comprises the steps of:
(a) identifying the the distribution of data contained in said second master distribution matrix corresponding to field values within the modified range:
(b) identifying the data records in the master distribution data file having field values within said modified range;
(c) patitioning a listbox with said third display;
and (d) displaying the field values of the records identified in step (b) within said listbox.