JPH04340623A - Data grouping processing method - Google Patents

Abstract

(57) [Summary] This bulletin contains application data before electronic filing, so abstract data is not recorded.

Description

[Detailed description of the invention]

0001

[Industrial Application Field] The present invention is directed to a given record.
The present invention relates to a data grouping processing method for grouping common data regarding desired items within a data group.

0002

2. Description of the Related Art FIG. 11 is a diagram illustrating data grouping processing. Assume that there is record data such as record data group 1 shown in Figure 11, in which individual name, age, and place of birth are linked, and the number of people by place of birth is to be investigated with respect to these record data. . Note that numeral 1 in the figure is a record data group, and 2 is an output result.

[0003] In this specification, the process called grouping process means a process for performing at least the following two processes. That is, one of them is a process (generally called grouping) in which individuals are grouped according to their place of birth in order to find the average age for each place of birth. Another method is to combine the same birthplaces into one in order to investigate the distribution of birthplaces (
(generally called distinct processing). In any case, these processes are called grouping processes in this specification because record data are grouped using, for example, place of birth as a common item.

FIG. 12 shows a processing flow of conventional grouping processing. (S1): In step S1, data in the given record data group is read. (S2): In step S2, the read data is sorted. At this time, sorting processing is performed on all target data. (S3): In step S3, grouping processing is performed.

FIGS. 13 to 16 show conventional processing modes. The processing mode is shown as obtaining processing results as shown in FIG. In the initial string creation process shown in Figure 13, the record string shown in Figure 11 is
Examine the record data one by one from data group 1 and edit it into a record with only the place of birth selected ((1)). Records are stored in the data pool 3-1 (assuming that six records are stored in one data pool), sorted within the data pool 3-1, and the initial string 4-1 is created. is written onto the disk memory ((2)). Similarly, the initial strings 4-2, 4-3, .
．． ．． Obtain ((3), (4), (5)).

In the intermediate merging process shown in FIG. 14, initial strings 4-1 and 4-2 are combined, the initial strings are merged, and intermediate string 5-2 is merged.
Obtain 1 ((1)). Similarly, the intermediate string 5-
Get 2. In the final merging process shown in FIG. 15, intermediate strings 5-1 and 5-2 are merged to obtain final string 6.

[0007] In the process of duplicate exclusion and count calculation shown in FIG. 16, data is extracted from the final string 6 and edited into a record 7 containing the place of birth and the number of people. Add (+1) the number of people ((1)). Then, when a person from a different birthplace appears, output it to the record for that birthplace ((2)
). Note that the reference numeral 2 shown in FIG. 16 is the same as the output result 2 shown in FIG.

[0008]

[Problems to be Solved by the Invention] As mentioned above, in the conventional case, after sorting all the data in a given record data group and obtaining the final string 6,
As shown in 6, output result 2 is obtained. For this reason, a large amount of processing time is required for sorting when obtaining the final string 6.

The object of the present invention is to reduce the time spent on sorting as much as possible.

0010

[Means for Solving the Problems] FIG. 1 shows a diagram of the basic configuration of the present invention. Reference numeral 101 in the figure represents an initial string generation process, and 102 represents a merge process. (S4): In step S4, individual data is read from the given record data group 1. (S5): In step S5, the read data and the already read data are sorted and grouped. (S6): In step S6, when step S5 is performed for a predetermined number of data, that amount is written on the disk memory as an initial string, and step S5 is performed again from the next data. (S7): Initial string generation processing step 10 as described above.
1 is completed and a plurality of initial strings are obtained, the initial strings are merged to obtain the final string. (S8): Grouping processing is performed using the final string.

[0011]

[Operation] In the case of the present invention, in step S5 shown in the figure, sorting is performed for each predetermined number of data, and grouping processing is performed. As a result, the obtained initial string is in a state as a result of grouping processing. This saves processing time during subsequent merging and grouping processing.

0012

Embodiment FIG. 2 is a diagram illustrating the initial string generation process. Reference numerals 1, 4-i, and 101 in the figure correspond to those shown in FIG. 11 and FIG. Given record
Data are extracted and input one by one from data group 1. Then, the data is sorted with the data in the data pool 11 that has already been input, and grouping processing is performed. A predetermined number of data resulting from the sorting and grouping processing is written on the disk as an initial string 4-i.

FIG. 3 is a diagram illustrating the intermediate merging process. The code 4-i in the figure is the initial string, and 5-i is the initial string.
i is an intermediate string. The initial strings 4-i and 4-(i+1) obtained in FIG. 2 are extracted, set in the string input area 12, and merged. The result is written onto disk memory as an intermediate string 5-p. Similarly, intermediate string 5-(
p+1). ．． ．． is obtained, but the obtained intermediate strings 5-i and 5-(i+1) are similarly merged to reduce the number of strings to a desired number.

FIG. 4 is a diagram illustrating the final merging process. The symbols in the figure correspond to those in FIG. In FIG. 4, the process continues until the number of strings is reduced to, for example, one. 5 to 7 show processing modes in the case of the present invention. The processing mode is shown as obtaining processing results as shown in FIG.

In the initial string generation process shown in FIG. 5, record data is examined one by one from the record data group 1 shown in FIG. )). Data pool 3-
1 and store the records in the data pool 3-1.
Sort within, perform grouping processing, and write initial string 4-1 to disk memory ((2))
. Similarly, initial strings 4-2, 4-3,
Obtain 4-4 ((3), (4), (5)).

In the intermediate merging process shown in FIG. 6, initial strings 4-1 and 4-2 are combined, the initial strings are merged, and the intermediate strings 5-1 and 4-2 are merged.
((1)). Similarly, intermediate string 5-2 is obtained ((2)). In the final merging process shown in FIG. 7, intermediate strings 5-1 and 5-2 are combined and merged between the intermediate strings to obtain output result 2.

As described above, in the present invention, processing time is shortened by appropriately performing necessary grouping processing even in the middle of processing. Note that, as shown in FIGS. 5, 6, and 7, the following two methods can be adopted for sorting and grouping processing. One method is to store data in the data pool 3-i and then sort the input data with the already input data to perform duplication exclusion, as shown in Figure 5. . Another method is to use tournament sorting.

FIGS. 8, 9, and 10 show processing modes of tournament sorting. In the figure, the record value is "3"
, "4", "5", "7", "8", and "9". As shown in Figure 8, record values “4”, “5”
”, “6”, “8”, and “9” are taken first to obtain the minimum value record by tournament. In this case, records with the value "4" are extracted. In this extracted case, as shown in the white circle in the figure,
A blank space is given to the location where the extracted record existed.

Assume that a record with a value of "7" is input as shown in FIG. 9 under a state in which a record with a value of "4" is extracted in FIG. In this case, since the value "7" is weaker than the value "4" extracted above, a record with the value "7" is set in the blank leaf as described above, and a tournament is held. . In the case of Figure 9, the value “
The record with "5" will be the winning record and will be extracted in the same way as in the case of FIG.

Assume that a record with a value of "3" is input as shown in FIG. 10 under a state in which a record with a value of "4" is extracted in FIG. In this case, the value "3"
is stronger than the above-extracted value “4”, so the value “3” is stronger than the value “4” extracted above.
A record with "" is flagged and set in the leaf section, and the tournament is held excluding the record with the flag. In the case of FIG. 10, the record with the value "5" is the winning record. Note that if the situation shown in FIG. 10 continues several times, all records existing in the leaf section will be flagged. When this situation occurs, each record extracted by the previous tournament is combined into one string. Then, the flags are removed from all records to which the above flags have been attached, and the tournament is started. In other words, processing to obtain a new string begins.

[0021] In Fig. 8, when a record with the value "4" is extracted, a record with the same value as the record with the value "4" and any record existing in the leaf section is input. If so, the input record will only be counted, and the next record will be input and set in the blank areas shown in FIGS. 8, 9, and 10.

[0022]

As described above, according to the present invention, grouping processing is performed as appropriate even in the middle of processing, and processing targets are grouped together, thereby shortening the overall processing time.

[Brief explanation of drawings]

FIG. 1 is a diagram showing the principle configuration of the present invention.

FIG. 2 is a diagram illustrating an initial string generation process.

FIG. 3 is a diagram illustrating an intermediate merge processing process.

FIG. 4 is a diagram illustrating a final merge processing process.

FIG. 5 shows a processing mode in the case of the present invention.

FIG. 6 shows a processing mode in the case of the present invention.

FIG. 7 shows a processing mode in the case of the present invention.

FIG. 8 shows a processing mode of tournament sorting.

FIG. 9 shows a processing mode of tournament sorting.

FIG. 10 shows a processing mode of tournament sorting.

FIG. 11 is a diagram illustrating data grouping processing.

FIG. 12 shows a flow of grouping processing in the conventional case.

FIG. 13 shows a conventional processing mode.

FIG. 14 shows a conventional processing mode.

FIG. 15 shows a conventional processing mode.

FIG. 16 shows a conventional processing mode.

[Explanation of symbols]

101 Initial string generation process 102 Merge process 1 Record data group 2 Output result 3 Data pool 4 Initial string 5 Intermediate string 6 Final string

Claims (2)

[Claims] [Claim 1] A data grouping processing method that performs processing to group together data having common items within the record data group for a given record data group, wherein: After extracting multiple pieces of data, data that has common items among the multiple pieces of data are combined into an initial string (4).
Initial string generation process (101) that sequentially generates
and between the plurality of initial strings (4) generated in the initial string generation process (101),
The configuration is configured to sequentially execute a merging process (102) of extracting data and combining data having common items, and grouping data having common items within the given record data group. A data grouping processing method characterized by the following. Claim 2: The initial string generation process (1)
In 01), when generating multiple initial strings (4), the data in the record data group given above is arbitrarily divided into multiple groups, and within the group, the above common items are 2. The data grouping processing method according to claim 1, wherein data having the following information is grouped together.