JPH06215028A - Multi-level summary data generating system - Google Patents

Abstract

PURPOSE:To accelerate processing by reducing arithmetic at the time of multihierarchical summary data generating arithmetic. CONSTITUTION:Data with symbolizing key to identify a node number are stored in a data storage means 107, concerning data inputted by a data input means 103, the master node numbers of previous data and present input data are acquired from a node information input means 102 by a summary arithmetic means 102, when those numbers are equal, the arithmetic of arithmetical rules is performed and when they are different, the arithmetic result and new data with symbolizing key added the master node number of previous data are passed through a symbolizing key applying means 105 and outputted from a data output means 106 to the data storage means.

Description

Detailed Description of the Invention

[0001]

BACKGROUND OF THE INVENTION 1. Field of the Invention The present invention relates to a method of generating summary data, and more particularly to a method of obtaining summary data of each node for multi-layer data.

[0002]

2. Description of the Related Art Conventionally, when this kind of summary record is generated, a key is set so that nodes can be classified for each hierarchy, all lower-order summary data is obtained according to the key, and next higher-order summary data is obtained. When seeking
The calculation method using the lower-level summary data obtained previously as input data was repeated for the number of layers.

An example of such a system is disclosed in Japanese Patent Laid-Open No. 1-1.
97882, "Region Calculation Method".

FIG. 7 is an area division diagram of a conventional area calculation method, which is an application example in the case of creating shape data to be the object of mechanical design by CAD. The decomposed shape A as indicated by the diagonal lines in FIG. 7A is divided into a plurality of regions to form a tree-structured data structure with a hierarchical representation, and the set operation of regions where speed is a problem is accelerated in parallel. Is what you are trying to do.

In FIG. 7 (a), the decomposed shape A is represented as a tree-like representation in a two-dimensional arrangement.
It is divided into four into 04, and the part 301 is further divided into 301-1 and 30.
It is divided into 1-2, 301-3, 301-4 ...
304-1-1, 304-1-2, 304-1-
3, 304-1-4 ..., and is divided as shown in FIG.
As shown in FIG. 7, if each of the divided parts is included in the area of the shape A indicated by diagonal lines, it is indicated by ●, and if it is outside the area, it is indicated by ○, and the four divided parts in FIG. ,
In the order of the upper left and the upper right, recording is performed by associating from left to right in FIG.

The hierarchical expression data shown in FIG. 7 (b) also has a pointer and "0, 1" on the memory as shown in FIG. 8 (a).
It is represented by the numerical value of 0, 0 represents the vacancy of 1 and 1 represents the fruit of ●.

Next, this data is converted into a one-dimensional symbol string. This conversion traces the tree structure of FIG. 7 (b) or FIG. 8 (a) vertically from left to right, and when descending one level down, it is '(', ● for '1', ○ for '0'. The decomposition shape A is converted into a symbol string of A; ((1101) 10 ((1101) 101)) by describing in order, "," when the hierarchy goes up one step, """.

Similarly, the decomposed shape B shown in FIG. 8B can be converted into a symbol string of B; (0 (0001) (0001) 1).

In order to obtain the shape region C = (A and B) shown in FIG. 8C from "A, B", the symbol string "A, B" is sequentially taken out from the beginning, and the set operation is partially performed. Performed as a series of operations, C = (A and B)
Is obtained as C; (0 (0001) 0 ((1101) 101)).

As described above, 30'-1, 3'are used as the key symbols for identifying the nodes in the hierarchy with '(', ')'.
01-2, 301-3, 301-4, or 304
-1-1, 304-1-2, 304-1-3, 304-
Higher-order summary data is obtained from lower-order summary data such as 1-4, and the operation is repeated. Although CAD is given here as an example, the invention is not limited to CAD.

[0011]

In such a conventional method, in order to obtain the higher-order summary data, all the lower-order summary data are obtained and output, and the obtained summary data is input to perform the summary calculation. However, there was a problem that the input and output of data increased and the processing took time.

In view of the above-mentioned problems, the present invention stores information expressing the interrelationship of nodes constituting a multi-layer, thereby reducing the data input / output in the multi-layer summary data generation and increasing the speed. The purpose is to provide a multi-level summary data generation method that enables the.

[0013]

According to the multi-layer summary data generation method of the present invention, individual raw data are classified, and the data obtained by performing a summary operation on the same group after classification is used as summary data and classified. In the method of generating multi-level summary data that configures a multi-level hierarchy by setting the group as an element of a higher-level group, at the beginning of the summary data obtained as raw data as input data and output data Data storing means for storing data with a symbolized key to which a symbolic key for identifying a node number is stored, and data for inputting raw data from the data storing means one by one and outputting EOF at the end of the input data Input means, command input means for specifying input file name for inputting data and output file name for outputting data, and symbol Data output means for inputting the data with symbolized key from the key giving means and outputting it to the data storage means, node information input means for inputting the node number of a certain node and returning the node number of the parent of the node, and summary computing means From the summary operation result input from the above and the parent node number of the previous data, a new symbolization key is created, and data with a symbolization key is newly added to the operation result and output to the data output means. The raw data is input from the data input means one by one until the EOF is returned, and the node numbers of the previous data and the current data that have been stored are output to the node information input means, and the node number of the parent of the node is input. If the node numbers of the parents of the two nodes are the same, the data is calculated according to the calculation rule. If they are not the same, the calculation is performed up to that point. The parent node number of the fruit before the data has a summary calculation means for outputting to the symbolized key providing means.

Further, in the summary calculation means, a level count representing the current level is counted, a summary calculation result storage area is prepared for each level, and when the comparison of the parent node numbers of the two nodes is not equal, the calculation is performed. The result is stored in the data storage means by the data output means after the key is added by the symbolized key giving means, and the level count is increased by one level and the storage area lower by one level is added to the storage area of the current level. It is characterized in that arithmetic processing is performed to store raw data in a storage area at a lower level.

Further, the EOF is assigned a node number of 0 which never matches the parent node number, and when the EOF is returned from the data input means, the operation result is added with a key by the encoding key giving means. The data output means stores the data in the data storage means, the level count is increased by one level, the storage area lower by one level is added to the storage area of the current level, and the data is stored when the level count reaches the highest level. It is characterized in that arithmetic processing for closing the storage means is performed.

[0016]

According to the above structure, information to which a symbol key for identifying the node number expressing the mutual relation of the nodes forming the multi-layer is added is stored in the data storage means, and the two nodes inputted from there are stored. Whether the parent nodes are the same is determined by the information of the node information input means, and the summary calculation means performs the summary calculation of the calculation rule when both the parent node numbers are the same, and when the different parent node numbers are different from the already calculated calculation result. The new symbolized keyed data to which the parent node number of the previous data is added is stored in the data storage means, the level count is increased by one level, and the storage area one level lower is added to the current level storage area. Since the raw data is stored in the storage area at the lower level of the level, the lower level summary data as input data is retained after being output, and it is used to store the upper level summary data. Data can be calculated.

[0017]

An embodiment of the present invention will be described in detail with reference to the drawings.

FIG. 1 is a block diagram of a multi-layer summary data generation system according to an embodiment of the present invention. Referring to FIG.
One embodiment of the present invention is a command input unit 101, a node information input unit 102, a data input unit 103, a summary calculation unit 104, a symbol key addition unit 105, a data output unit 106, and a data storage unit 107. Composed of and.

The command input means 101 is means for inputting an input file name for inputting raw data and an output file name for outputting data with symbolized keys, and the node information input means 102 is for a certain node. The data input means 103 is a means for inputting a node number and returning the parent node number of the node.
The data input from the input file on the data storage means 107 designated by the above is calculated one by one in the summary calculation means 104.
To the end of the input data and EOF (End of of
Is a means for outputting File), and is a summary calculation means 1
04 inputs data one by one from the data input means 103 until EOF is returned, outputs the stored previous data and the node number of the current data to the node information input means 102,
Enter the node number of the parent of the node, and when the node numbers of the parents are equal, perform a summary operation on the data according to the operation rules, and if they are not equal, use the symbol to indicate the operation result of the summary operation up to that point and the parent node number of the previous data. Key giving means 1
It is a means for repeating the above operation until the data input means 103 returns EOF from the data input means 103, and the encoding key addition means 105 uses the summary operation result input from the summary operation means 104 and the parent node number of the previous data. Is a means for creating a new symbol key, adding a new symbol key to the operation result, and outputting the created data with the symbol key to the data output means 106. The data output means 106 is the symbol key adding means. 105 is a means for inputting the data with symbolized key from 105 and outputting the data with symbolized key to the output file on the data storage means 107 designated by the command input means 101. Store the raw data and the data with symbolization key with the symbolization key for identifying the node number added to the beginning of the obtained summary data. Is a place means.

Next, the flow of the entire processing will be described.

An input file name and an output file name are input from the command input means 101, and the summary calculation means 104
Inputs the raw data from the data input means 103 one by one until the EOF is returned, and outputs the node numbers of the stored previous data and current data to the node information input means 102,
Enter the parent node number of each node, compare the node numbers of each parent for equality, and if they are equal, calculate the data according to the calculation rules.If they are not equal, calculate the result of the summary calculation up to that time and the previous data The parent node number is output to the symbolic key assigning means 105, and the symbolic key assigning means 105 outputs the symbolized key-attached data obtained by adding the symbolized key to the calculation result to the data output means 106, and the data output means. 106 receives the symbolized key-attached data from the key giving means 105 and outputs it to the data storage means 107,
When EOF is returned from the data input means 103, the summary calculation means 104 outputs the calculated summary data.

Next, the operation outline of the main processing will be described.

FIG. 2 is a flow chart showing an outline of the operation of the summary calculation means 104 of FIG. 1. The level count used in the processing shown in FIG. 2 represents the current level, and the summary calculation result area is for each level. It shall be prepared.

First, the area for storing the result of the summary calculation is initialized for all levels, the level counter is initialized (step 201), and the input file name and the output file name are input from the command input means 101 (step 20).
2) Open the designated input file and output file on the data storage means 107 (step 203), input one piece of raw data from the data input means 103, and set the level count to the lowest level (step 204). When EOF is returned from the data input means 103, step 2
If it is not EOF, the process jumps to step 206 (step 205), the node number of the input data is stored (step 206), and the focused node number of the input data and the data stored immediately before is stored in the node information. The information is output to the input unit 102, the node numbers of the respective parents are acquired (steps 207 and 208), and the node information input unit 10
The node number of the parent of the current data input from 2 and the node number of the parent of the previous data input are compared, and if they are equal, the process jumps to step 210, and if they are not equal, the process jumps to step 211 (step 209). , The data having the same parent is calculated by the summary calculating means 104, and the process jumps to step 204 (step 210). If they are not equal, the calculation result calculated by the summary calculating means 104 and the parent node of the data previously input A new symbolized key-attached data is created from the number by the symbolized key giving means 105, stored in the data storage means 107 by the data output means 106 (step 211), and the level of interest is increased by one level (step 212). Add the storage area one level lower to the storage area of the current level (step 213), one level When the raw data is stored in the storage area of the order and jumps to step 207 (step 214) and EOF is returned from the data input means 103 in step 205, the calculation result calculated by the summary calculation means 104 and the previous one are input. A new symbolized key-attached data is created by the symbolized key assigning means 105 from the parent node number of the data, stored in the data storage means 107 by the data output means 106 (step 215), and the level of interest is increased by one level. (Step 216) Add the storage area of one level lower to the storage area of the current level (step 217), determine whether the level counter is the highest level, and if it is the highest level, jump to step 219.
Otherwise, jump to step 215 (step 21
8) If the level counter is at the highest level, the data storage means 107 is closed and the process ends (step 21).
9).

Next, specific processing will be described.

The command input means 101, the node information input means 102, the data input means 103, the summary calculation means 104, the symbol key addition means 105, and the data output means 106 are computer programs, and the respective command input programs are used. , Node information input program, data input program, summary calculation program, symbol key addition program, data output program,
The data storage means 107 is a file.

The number of layers of the multi-level summary data is 3, and
An example having a tree structure of nodes such as

In the command input, an input file name and an output file name are designated.

In FIG. 3, “Nationwide” with node number 1 is “Kanto” with node number 2, “Tokyo” with node number 3, “Kanagawa” with node number 4, and “Chubu” with node number 5, Consider a regional axis composed of node number 6 “Aichi” and node number 7 “Shizuoka”.

In the above node configuration, the input data is
"Tokyo 10", "Kanagawa 20", "Aichi 30",
There are 4 records of "Shizuoka 40". (FIG. 5) Further, the summary data is composed of a node number and a record part as shown in FIG.

When the record "Tokyo 10" is input, the node number 3 is stored first, and then the data portion "10" of the record is stored in the 2-level operation area by the summary operation program.

When the record "Kanagawa / 20" is input, the node number 4 is stored, the node number 4 of the current record "Kanagawa" is output to the node information input program, and the node of the current node "Kanto" is output. Enter the number 2. The summary calculation program determines whether the parent nodes are the same. Since the node number of the parent node is equal to 2, the previous data “1” is set in the 2-level operation area of the summary operation program.
0 "and the current data" 20 "are calculated.

Next, when "Aichi.30" is input, the node number 6 is stored, the node number 6 of the current record "Aichi" is output to the node information input program, and the node number 5 of the parent node "Chubu" is output. Enter. The summary calculation program determines whether the parent nodes are equal. Since the node number of the parent node is 5, which is different from the parent node of the previous data, a new symbol key is added to the summary data “30” calculated by the symbol calculation key addition program by the summary calculation program by the region axis “Tokyo”. In the parent node "Kanto" of "" and "Kanagawa", add "Kanto30" (2), output to a file by the data output program, and output the summary data "30" to the one-level operation area. Add in. Next level up by one and the current record is "Chubu"
As the node number 5 in the node information input program
Is output, and the node number 1 of the parent node “national” is input. The summary calculation program determines whether the parent nodes are equal. Since the node number of the parent node is 1 and equal,
Hold the 1-level operation area of the summary operation program,
The 2-level calculation area is cleared. After that, the current data "Aichi-3
The data part "30" of 0 "is output.

Next, when "Shizuoka / 40" is input, the node number 7 is stored, the node number 7 of the current record "Shizuoka" is output to the node information input program, and the parent node "Chubu" node is output. Enter the number 5. The summary calculation program determines whether the parent nodes are equal. Since the parent nodes are equal to 5, the operation is performed with the previous data “30”. As a result, "70" is output to the 2-level operation area of the summary operation program.

Next, when EOF is returned, the input file is closed, and since EOF is assigned 0 as the node number, the parent node will never match the conventional one. A new symbol key is added to the data "70" calculated by the symbol key addition program at the parent node "Chubu" of the regional axes "Aichi" and "Shizuoka".
"Chubu-70" (5) is added, and the data output program outputs the data to a file and adds the calculated summary data "70" to the one-level operation area. Next, increase the level by one, and by the encoding key addition program,
A new symbol key is added to the calculated data "100" at the parent node "national" where the regional axes are "Kanto" and "Chubu", and "national 100" (1) is added to the file by the data output program. Output. (FIG. 6) Through the above processing, summary data of combinations of all nodes is generated. As described above, in the present embodiment, as compared with the conventional example, it is possible to reduce the calculation of the summary data based on the related information between the multi-layer nodes and to achieve the speedup.

[0036]

As described above, according to the present invention, information expressing the mutual relation of the nodes forming the multi-layer is stored in the data storage means, and whether the upper two nodes are the same as the input two nodes. Whether or not the higher-order node of the previously input data and the currently input data is the same is determined by the summary operation means based on the information of the node information input means, and if the same, the higher-order summary operation is performed. When different, the already-calculated summary calculation result is output, so that there is an effect that data input / output can be reduced and the speed can be increased.

[Brief description of drawings]

FIG. 1 is a configuration diagram of a multi-layer summary data generation method according to an embodiment of the present invention.

FIG. 2 is a flowchart of processing of a summary calculation means shown in FIG.

FIG. 3 is a diagram showing an example of a node structure of summary data according to the present invention.

FIG. 4 is a diagram showing an example of a structure of summary data of the present invention.

5 is a diagram showing contents of keys of the summary data shown in FIG.

FIG. 6 is a diagram showing summary data created by the keys shown in FIG.

FIG. 7 is an area division diagram of a conventional area calculation method, (a)
Shows a decomposed shape, and (b) shows a tree structure.

FIG. 8 is an explanatory diagram of a conventional area calculation, in which (a) shows a memory structure, and (b) and (c) show decomposed shapes.

[Explanation of symbols]

 101 Command Input Means 102 Node Information Input Means 103 Data Input Means 104 Summary Computation Means 105 Symbolization Key Adding Means 106 Data Output Means 107 Data Storage Means

Claims (3)

[Claims] 1. The data obtained by classifying individual raw data and performing a summary operation on the same group after classification is used as summary data, and the classified group becomes a higher-order group element and is then made into a large number. In the method to generate multi-level summary data that configures the hierarchy of levels, the raw data as input data and the summary data obtained as output data are symbolized by adding a symbol key to identify the node number at the beginning. Data storage means for storing keyed data, and raw data is input from the data storage means one by one,
Data input means for outputting EOF at the end of input data, command input means for designating an input file name for inputting data and an output file name for outputting data, and a symbol key attaching means for attaching a symbol key Output data output means for inputting data to the data storage means, node information input means for inputting the node number of a certain node and returning the node number of the parent of the node, summary operation result and previous data input from the summary operation means A new symbolic key is created from the parent node number of, the symbolic key attaching means for newly adding the symbolic keyed data to the operation result and outputting the data to the data output means, and the raw data from the data input means 1 to return
When the node numbers of the previous data and the current data that have been stored are output to the node information input means and the parent node number of the node is input and the node numbers of the respective parents of the two nodes are equal Has a summary calculation means for performing a summary calculation on the data according to the calculation rule, and outputting the calculation result obtained by the summary calculation until that time and the node number of the parent of the previous data to the encoding key giving means when the data are not equal. Multi-level summary data generation method. 2. The summary calculation means counts the level representing the current level, prepares a summary calculation result storage area on the data storage means for each level, and compares the parent node numbers of two nodes equally. If not, the operation result is stored in the data storage means by the data output means after adding the key by the symbolization key provision means, and the level count is increased by one level and the storage area lower by one level is added to the storage area of the current level. 2. The multi-layer summary data generation method according to claim 1, wherein the arithmetic processing is performed to store the raw data in a storage area lower by one level. 3. A node number of 0 which never matches the parent node number is assigned to the EOF, and when the EOF is returned from the data input means, a key is added to the calculation result by the encoding key giving means. The data storage means stores the data in the data storage means, increases the level count by one level, adds a storage area of one level lower to the storage area of the current level, and outputs the data storage means when the level count reaches the highest level. 3. The multi-layer summary data generation method according to claim 1 or 2, characterized in that a calculation process for closing is performed.