JP2023160301A - data compression device - Google Patents

Abstract

To solve the problem in which: occurrence of a shift in every data acquisition timing reduces compressibility.SOLUTION: A data compression device comprises interpolation means, compression means, and data storage means. The interpolation means, from the time series of history data including time data and statistic data associated with the time data and data indicating a predetermined collection interval, calculates, by interpolation calculation, the time series of post-interpolation history data including post-interpolation time data and post-interpolation statistic data associated with the post-interpolation time data. The compression means calculates compression history data including post-interpolation statistic data extracted from the time series of the post-interpolation history data in order of the time indicated by the post-interpolation time data. The data storage means outputs the compression history data to a database.SELECTED DRAWING: Figure 1

Description

The present invention relates to a data compression device, a data compression method, and a program.

2. Description of the Related Art A data compression device is known for reducing the amount of information that associates data with the acquisition time at which the data was acquired. For example, the technology described in Patent Document 1 (hereinafter referred to as the first related technology) includes a plurality of data acquired at predetermined sampling intervals from an acquisition start time and an acquisition time of each of the plurality of data. A plurality of pieces of data are extracted from the compression target information in the order of acquisition time, and compressed information is output including the plurality of data extracted in the order of acquisition time, the acquisition start time, and interval information indicating the sampling interval. .

JP2017-022522A JP 2021-052263 Publication Patent No. 4911061

The first related technique functions effectively in an environment where there is no deviation in the timing of data acquisition each time. However, in an environment where there is a shift in the timing of data acquisition each time, the precondition that "a plurality of data are acquired at predetermined sampling intervals from the acquisition start time" does not hold, and the compression ratio decreases.

An object of the present invention is to provide a data compression device that solves the above-mentioned problems.

A data compression device according to one embodiment of the present invention includes:
A time series of historical data including time data and statistical data associated with the time data, and data indicating a predetermined collection interval are correlated with the post-interpolated time data and the post-interpolated time data by an interpolation operation. interpolation means for calculating a time series of interpolated history data including post-interpolated statistical data;
Compression means for calculating compressed history data including the post-interpolated statistical data extracted from the time series of the post-interpolated history data in the order of time indicated by the post-interpolated time data;
data storage means for outputting the compression history data to a database;
It is configured to include.

Further, a data compression method according to another embodiment of the present invention includes:
A time series of historical data including time data and statistical data associated with the time data, and data indicating a predetermined collection interval are correlated with the post-interpolated time data and the post-interpolated time data by an interpolation operation. Calculate the time series of the interpolated historical data including the interpolated statistical data,
calculating compressed history data including the post-interpolated statistical data extracted from the time series of the post-interpolated history data in the order of time indicated by the post-interpolated time data;
outputting the compression history data to a database;
It is configured as follows.

Further, a program according to another embodiment of the present invention is
to the computer,
A time series of historical data including time data and statistical data associated with the time data, and data indicating a predetermined collection interval are correlated with the post-interpolated time data and the post-interpolated time data by an interpolation operation. a process of calculating a time series of interpolated historical data including post-interpolated statistical data;
a process of calculating compressed history data including the post-interpolated statistical data extracted from the time series of the post-interpolated history data in the order of time indicated by the post-interpolated time data;
a process of outputting the compression history data to a database;
is configured to perform the following.

By having the above-described configuration, the present invention can achieve a high compression ratio even in a situation where a shift occurs in the timing of data acquisition each time.

FIG. 1 is a block diagram of a data compression device according to a first embodiment of the present invention. FIG. 7 is a diagram illustrating a specific example of a data structure and a configuration of managed objects in the second related technology. 3 is a diagram showing an image of the registered contents of the fact table 8 in FIG. 2 (uncompressed). FIG. It is a figure showing an example of composition of a management system concerning a 2nd embodiment of the present invention. It is a figure which shows the example of registration of statistical data in the management system based on the 2nd Embodiment of this invention. FIG. 6 is an explanatory diagram of linear interpolation in the management system according to the second embodiment of the present invention.

Next, embodiments of the present invention will be described in detail with reference to the drawings.
[First embodiment]
FIG. 1 is a block diagram of a data compression device 100 according to a first embodiment of the present invention. Referring to FIG. 1, a data compression device 100 includes an interpolation means 101, a compression means 102, and a data storage means 103.

The interpolation means 101 inputs a time series of historical data and a predetermined collection interval. Each historical data includes time data and statistical data associated with the time data. The historical data is collected at the above-mentioned collection interval by a data acquisition means (not shown) as historical data including, for example, statistical data indicating the status of a managed device such as a server computer and time data indicating the time when the statistical data was acquired. It may be periodically acquired based on the data. However, there may be a shift in the timing of obtaining statistical data each time. For example, even if a management system is designed to collect the status of a managed device (for example, the load factor of a server computer) at a collection interval of 10 seconds, it is necessary to always accurately obtain the load factor every 10 seconds. There is no guarantee that the load factor will be obtained, and in some cases, the load factor may be obtained 10.1 seconds after the previous acquisition time, or the load factor may be obtained 9.9 seconds after the previous acquisition time.

Further, the collection interval may be time information recorded in advance on a recording medium accessible from the interpolation means 101. Alternatively, the collection interval may be time information statistically calculated by the interpolation means 101 as an average value of acquisition cycles from time data included in a time series of historical data.

Further, the interpolation unit 101 performs an interpolation operation on the time series of the input history data and the collection interval to calculate the time of the post-interpolated history data including the post-interpolated time data and the post-interpolated statistical data associated with the post-interpolated time data. Calculate the series. For example, the interpolation means 101 configures time data and statistical data t ⁿ and v ⁿ that constitute the n-th history data in the time series of the history data, and configures the n-th post-interpolated history data in the time series of the interpolated history data. When the interpolated time data and interpolated statistical data are t ^n' and v ^n' and the collection interval is C, the linearity given by equations (1) to (3) described later in the second embodiment and later is An interpolation calculation may be performed to calculate the time series of the interpolated history data. However, the interpolation calculation is not limited to linear interpolation, and may be Lagrangian interpolation or spline interpolation. Note that if post-interpolation data does not exist, such as the first data, pre-interpolation data is used.

The compression means 102 calculates compressed history data including post-interpolation statistical data extracted from the time series of the post-interpolation history data calculated by the interpolation means 101 in the order of time indicated by the post-interpolation time data. Further, the compression means 102 may further perform an XOR operation (eXclusive OR operation) on two pieces of post-interpolated statistical data in floating point format to compress the post-interpolated statistical data in the compression history data calculated above. good.

The data storage means 103 outputs the compression history data calculated by the compression means 102 to the database. Further, the data storage unit 103 may output to the database data in which information indicating the presence or absence of compression and information indicating the number of interpolated statistical data included in the compression history data are stored in an area for storing compression history data. The structure of the database is arbitrary. For example, the database may have a star schema structure. Further, the database may previously hold data indicating the collection interval in association with the compression history data. Alternatively, the data storage means 103 may output data indicating the collection interval to the database in association with the compression history data. Further, the data storage unit 103 may output time data, which is the starting point of the compression history data, to the database in association with the compression history data.

Further, the data compression device 100 may include a data decompression means (not shown). The data decompression means reads compressed history data from the database, decompresses it, and outputs interpolated history data.

The above-mentioned interpolation means 101, compression means 102, data storage means 103, and the above-mentioned data acquisition means and data decompression means are, for example, a computer having at least a processor and a memory, and a program executed on this computer. It can be realized.

Next, the operation of the data compression device 100 configured as above will be explained.

First, the interpolation unit 101 performs an interpolation calculation from a time series of historical data including time data acquired by a data acquisition unit (not shown) and statistical data associated with the time data, and data indicating a predetermined collection interval. A time series of post-interpolated history data including post-interpolated time data and post-interpolated statistical data associated with the post-interpolated time data is calculated.

For example, consider the following time series of historical data 1 to 6.
Historical data 1: Time April 15, 2022 09:00:00, statistical data 000
Historical data 2: Time April 15, 2022 09:00:11, statistical data 110
Historical data 3: Time April 15, 2022 09:00:19, statistical data 190
Historical data 4: Time April 15, 2022 09:00:31, statistical data 310
Historical data 5: Time April 15, 2022 09:00:40, statistical data 400
Historical data 6: Time April 15, 2022 09:00:51, statistical data 510

If the collection interval is 10 seconds, the interpolation means 101 calculates the following time series of interpolated history data 1' to 6' from the history data 1 to 6, for example, by linear interpolation.
Historical data 1': Time April 15, 2022 09:00:00, statistical data 000
Historical data 2': Time April 15, 2022 09:00:10, statistical data 100
Historical data 3': Time April 15, 2022 09:00:20, statistical data 200
Historical data 4': Time April 15, 2022 09:00:30, statistical data 300
Historical data 5': Time April 15, 2022 09:00:40, statistical data 400
Historical data 6': Time April 15, 2022 09:00:50, statistical data 500

Next, the compression means 102 calculates compressed history data including post-interpolation statistical data extracted from the time series of the post-interpolation history data calculated by the interpolation means 101 in the order of time indicated by the post-interpolation time data. For example, in the case of the time series of post-interpolated history data 2' to 6' as described above, the compression means 102 calculates compressed history data including the following six pieces of post-interpolated statistical data.
Statistical data after interpolation 000
Statistical data after interpolation 100
Statistical data after interpolation 200
Statistical data after interpolation 300
Statistical data after interpolation 400
Statistical data after interpolation 500

Next, the data storage means 103 outputs the compression history data calculated by the compression means 102 to the database. Further, the data storage means 103 outputs the time data of the history data 1 to the database in association with the compressed history data.

According to the data compression apparatus 100 configured and operated as described above, a high compression ratio can be achieved even in a situation where a shift occurs in the timing of data acquisition each time. The reason is that the data compression device 100 calculates a time series of interpolated history data at equal intervals in time from the time series of history data and data indicating a predetermined collection interval, and This is to calculate compression history data from the time series and output this compression history data to the database.

[Second embodiment]
Next, a second embodiment of the present invention will be described in detail.
[Features of this embodiment]
In this embodiment, the data compression method performed when the operation management system saves data collected from managed objects is improved. In the technology described in Patent Document 2 (hereinafter referred to as the second related technology), the collected data is compressed and stored using a method tailored to the two types of data, statistical data and time data. However, in this embodiment, a higher compression ratio is achieved by compressing and saving only the statistical data calculated by interpolation and not saving the time data.

[Problems to be solved by this embodiment]
The second related technology and the technology described in Patent Document 3 (hereinafter referred to as the third related technology) are technologies related to a mechanism and data structure for storing historical data on a management system that includes a server as a component. be.

The third related technology provides a method that uses a star schema data structure to make it easier to search historical information on statistical data for the same management target from multiple perspectives using multiple dimensional tables. are doing.

Further, the second related technology provides a method for compressing and storing time series time data and statistical data using compression methods such as double-delta compression and XOR compression based on the third related technology. FIG. 2 shows a specific example of a data structure and a configuration of managed objects in the second related technology. Further, FIG. 3 shows an image of the registered contents of the fact table 8 in FIG. 2 (uncompressed).

The second related technology provides a method for compressing and storing time series time data and statistical data using double-delta compression and XOR compression based on the third related technology. During operation, the data collection time varies slightly each time due to the influence of the system load. In this situation, double-delta compression, which is a time data compression method, does not have good compression efficiency.

For example, consider a system that collects data every 5 minutes as follows.
1:00:00.000
1:05:00.000
1:10:00.000

However, in reality, the collection intervals are as shown below, and are often not exactly 5 minutes apart.
1:00:00.005
1:05:00.100
1:10:00.079

In the above case, in the second related technique, the time data requires a data amount of 9 bits or more. In this case, the 4-byte data area can only compress two pieces of data at most.

[Solution means according to this embodiment]
In this embodiment, the problem is solved by the following method, which eliminates the need to store time data in the area where data is stored during compression.

First, target time data and statistical data are calculated as follows from information such as actually collected data and reference time.
(1) Time data: Add the collection interval to the reference time.
(2) Statistical data: Statistical data corresponding to the above time data is calculated from the reference time data and the time data and statistical data of the actually collected data using an interpolation method.

As a result of the above, time data at an arbitrary timing, that is, time data after interpolation, can be easily calculated from the reference time and collection interval information, and therefore is not stored in the database. Calculated from the reference time and collection interval information when acquiring data. Furthermore, the data area required for storing time data can be used as an area for storing statistical data.

[Configuration of this embodiment]
FIG. 4 is a diagram showing a configuration example of a management system according to this embodiment. The difference between the configuration example of the management system shown in FIG. 4 and the configuration shown in FIG. 2 is the data interpolation means 17 and the data cache 18 inside the thick line rectangle in FIG.

Generally, IT systems have servers, networks, etc. as system components. The management system in this embodiment has a configuration management component 1 that manages these components and has a role of holding and managing information on the components.

The configuration management component 1 stores information about the system to be managed in a configuration information DB (database) 3, and manages the information using the configuration management means 10. Further, the configuration management component 1 includes a dimensional table data addition unit 4 as a unit for reflecting configuration information in the history information DB 4.

The management system also includes a system monitoring component 2 as a function of monitoring the status of the system components. In the system monitoring component 2, the system information collection monitoring means 11 collects historical data of the target system. Further, the data collected by the system information collection and monitoring means 11 is transferred from the system monitoring component 2 to the history data addition means 5. The history data addition means 5 sends the data received from the system information collection and monitoring means 11 to the data interpolation means 17 of the data compression and decompression component 12.

The data compression/decompression component 12 uses the data interpolation means 17 to interpolate the history data passed from the history data addition means 5, and sends the interpolated data to the data compression means 13. The data interpolation means 17 uses a data cache 18 for temporarily storing data until the interpolation data can be sent to the data compression means 13. Information on collection intervals required for data interpolation is stored in advance in the configuration information DB3. The data interpolation means 17 acquires information on necessary collection intervals from the configuration information DB 3.

The data compression means 13 compresses the history data passed from the data interpolation means 17 and registers the compressed data in the fact table 8 of the history information DB 6. During compression, the data compression means 13 uses a data cache 15 that temporarily stores data until it can be registered in the database.

This operation operates periodically, and performance data and statistical data regarding the target system are periodically saved and accumulated as time-series historical data.

The history information DB 6 for storing history information of system management tools has a table structure generally called a star schema and is composed of a data table used as a fact table 8 and a plurality of dimensional tables 7 centered around the fact table. .

The fact table 8 includes statistical data registered as historical data of statistical information indicating the status of the system collected by the system information collection and monitoring means 11, time data indicating the date and time when the data was acquired, statistical data, etc. It consists of ID data associated with the table. The ID is used as an index for data on other dimension tables that are linked to data on the data table, and serves as a key indicating what kind of historical data is the statistical information registered in the data table.

The statistical data registered in the fact table 8 includes server statistical information such as server load information, event occurrence information, and the number of configuration changes. Further, the statistical data registered in the fact table 8 may include statistical information such as the number of servers. Tables other than fact table 8 store data indicating the configuration of the system to be managed.

The management system also includes report creation means 9 for displaying data registered in the history information DB 6 to users of the system. The report creation means 9 searches for and obtains data from the history information DB 6 from the data decompression means 14 of the data compression and decompression component 12, and creates a report.

[Operation of this embodiment]
The main operations of the management system according to this embodiment will be explained.

<Overview of data interpolation method>
The history data collected by the system information collection and monitoring means 11 in the system monitoring component 2 is sent from the history data addition means 5 to the data interpolation means 17 of the data compression and decompression component 12, and is temporarily stored in the data cache 18.

The data interpolation means 17 performs data interpolation using the following calculation formula.
t ^n' = t ^n-1' + C...(1)
v ^n' , = (1-α) x v ^n-1' + α x v ⁿ , ... (2)
α=(t ^n' - t ^n-1' )/(t ⁿ - ^{t n-1'} ) ... (3)
Here, t ^n' is time data after the n-th interpolation. v ^n' is statistical data after the n-th interpolation. t ⁿ is time data before the n-th interpolation. v ⁿ is the n-th pre-interpolation statistical data. C is the collection interval. FIG. 6 is an explanatory diagram of linear interpolation using the above equation. Using the interpolated data t ^n-1' , v ^n-1' at the previous collection timing and the data t ⁿ , v ⁿ actually collected at the current collection timing, interpolated data t ^n' , v ^n' are generated. calculate. Note that for t ^n-1 ′ and v ^n-1 ′, if post-interpolation data does not exist, such as the first data, pre-interpolation data is used.

<Overview of data storage method>
The history data collected by the system information collection and monitoring means 11 in the system monitoring component 2 is sent from the history data addition means 5 to the data interpolation means 17 of the data compression and decompression component 12, and is temporarily stored in the data cache 18.

The data interpolation means 17 performs the above-mentioned data interpolation based on the history data sent from the system information collection monitoring means 11 and the collection interval information acquired from the configuration information DB 3. The interpolated data is sent to the data compression means 13 and temporarily stored in the data cache 15.

The data compression means 13 stores the interpolated data in the history information DB 6 based on the following idea.

Columns that store statistical data stored floating point data. Focusing on the fact that normally, the statistical data to be saved is only positive values, in this embodiment, if the statistical data is a negative value, it is saved as an area to store the compressed data of the statistical data after the relevant time. do. Note that since the time data can be calculated from the reference time and the collection interval information acquired from the configuration information DB 3, it is not saved in the compressed data area.

A case will be described below in which a float type (32 bit) floating point value as shown in FIG. 5 is stored in the statistical data column. Based on the data information from the previous time, the statistical data is compressed and stored according to the following rules.

(1) 1st bit: Used to distinguish whether or not it is compressed data. A value of 1 means compression is ON, and a value of 0 means compression is OFF. As a floating point number, it represents the sign. If compression is performed, it is saved as a negative floating-point number. If it is positive, it is stored as normal statistical data.
(2) 2nd and subsequent bits: Stores statistical data and time data and statistical data of historical data after the later collection timing in the following format.
(2-1) Encoded data of the first statistical data. Generate and register using XOR compression. Note that the time data associated with the first statistical data is stored, for example, in the time data column of the fact table.
(2-2) The number of data to be stored after the second one is stored using 2 bits or 6 bits as shown below.
(2-2-1) When the number of stored data is 1 to 2, the 1st bit is set to 0, and the value of the remaining 1 bit represents 1 or 2 (0 for 1, 1 for 2).
(2-2-2) When the number of stored data is 3 or more, the first bit is set to 1 and the value of the remaining 5 bits represents the number of stored data.
(2-3) Store the encoded data of the second statistical data. Generate and register using XOR.
(2-4) Store the second and subsequent history data as much as possible in the same manner as the second data. Since the second and subsequent bits use a minimum of 1 bit, a maximum of 25 pieces of data can be stored.
(3) In the following cases, it is assumed that the data cannot be compressed and is stored in the normal way.
(3-1) When the number of data to be stored is not two or more (including cases where it is not possible to store data that satisfies the conditions of XOR compression)

<XOR compression>
Although it depends on the type, server statistical data often changes at similar values. Furthermore, XOR of two floating point numbers with similar values has a characteristic that many bits tend to become 0. For example, assume that there is time-series data of history data 1 to 3 having the following float type values.
(1) History data 1:
Time 1, statistical data 9.0 (binary representation is 01000001000100000000000000000000)
(2) History data 2:
Time 2, statistical data 13.0 (binary representation is 01000001010100000000000000000000)
(3) History data 3:
Time 3, statistical data 7.0 (binary representation is 01000000111000000000000000000000)

At this time, the XOR between the statistical data of historical data 1 and the statistical data of historical data 2 is
(4) 00000000010000000000000000000000
become.
Also, the XOR between the statistical data of historical data 2 and the statistical data of historical data 3 is
(5) 00000000101100000000000000000000
become.

At this time, the sequence of nine leading 0s and all the 0s following 1 can be encoded using the XOR (4). That is, the only significant bit of the XOR in (4) is "1". The significant bit of XOR in (5) is the "1011" part.

Utilizing such properties, the data compression means 13 generates data to be registered according to the following rules.
(1) If the XOR with the previous value is 0 (ie, the same value), write 0.
(2) If the XOR with the previous value is not 0, first write 1, and then write the following three pieces of information in order for the significant bits to be calculated.
(2-1) If the number of leading zeros can be written in 4-bit representation, write it.
(2-2) If the length of significant bits can be written in 3-bit representation, write it.
(2-3) Write the significant bit itself (up to 7 bits in length).
(3) If the conditions are not met, give up on storing the second and subsequent data. Furthermore, if the second and subsequent data cannot be stored, the data is registered using the normal method.

FIG. 5 shows an example of registering statistical data. When the value of the statistical data is 9.0, it is stored in the statistical data column as shown in the upper part of FIG. Furthermore, if compression is possible, the compressed data is stored as a negative floating-point value, as shown in the lower part of FIG.

<Effects of this embodiment>
As described above, according to the present embodiment, only the statistical data calculated by interpolation is compressed and stored, and time data is not stored, so that a higher compression ratio can be achieved. For example, in this embodiment, the 1, 9, 12, and 16 bit areas that were used to store time data in the second related technology can be used to store statistical data, so Data can be stored at a compressed rate. Specifically, compared to the second related technology, when storing float type (32 bit) floating point values, the maximum number of data storage can be increased from 14 to 25. Furthermore, in the case of 9 bits of statistical data and 9 bits of time data, the number can be increased from 2 to 3. Thus, according to this embodiment, it is possible to provide a more efficient data compression means without changing the simple star schema data structure using a relational database.

<Fields where this embodiment can be used>
This embodiment can be used in a system that periodically collects and accumulates time-series data of a managed object, such as a server operation management system. In addition, it is also considered to be effective for use in IOT devices other than computers.

Although the present invention has been described above with reference to the embodiments described above, the present invention is not limited to the embodiments described above. The configuration and details of the present invention may be modified in various ways within the scope of the present invention by those skilled in the art.

INDUSTRIAL APPLICATION This invention can be utilized for the system which periodically collects and accumulates the data showing the state etc. of a managed object in the management system which manages a server computer, IOT device, etc. as a managed object.

100 Data compression device 101 Interpolation means 102 Compression means 103 Data storage means

Claims (10)

A time series of historical data including time data and statistical data associated with the time data, and data indicating a predetermined collection interval are correlated with the post-interpolated time data and the post-interpolated time data by an interpolation operation. interpolation means for calculating a time series of interpolated history data including post-interpolated statistical data;
Compression means for calculating compressed history data including the post-interpolated statistical data extracted from the time series of the post-interpolated history data in the order of time indicated by the post-interpolated time data;
data storage means for outputting the compression history data to a database;
A data compression device comprising: The interpolation means configures time data and statistical data t ⁿ and v ⁿ that constitute the n-th history data in the time series of the history data, and constitutes the n-th interpolated history data in the time series of the interpolated history data. When the interpolated time data and interpolated statistical data are t ^n' and v ^n' and the collection interval is C,
t ^n' = t ^n-1' + C...(1)
v ^n' , = (1-α) x v ^n-1' + α x v ⁿ , ... (2)
α=(t ^n' - t ^n-1' )/(t ⁿ - ^{t n-1'} ) ... (3)
Calculating the time series of the interpolated history data by the following calculation,
The data compression device according to claim 1. The compression means further compresses the interpolated statistical data in the compressed history data by performing an XOR operation on two pieces of the interpolated statistical data in floating point format.
The data compression device according to claim 2. further comprising data decompression means for reading and decompressing the compressed history data from the database and outputting the interpolated history data;
The data compression device according to claim 1. The interpolation means obtains, as the time series of the history data, a time series of history data including statistical data indicating the state of the server and time data indicating the time at which the statistical data was obtained from at least one server; further comprising a data acquisition means for outputting to the
The data compression device according to claim 1. A time series of historical data including time data and statistical data associated with the time data, and data indicating a predetermined collection interval are correlated with the post-interpolated time data and the post-interpolated time data by an interpolation operation. Calculate the time series of the interpolated historical data including the interpolated statistical data,
calculating compressed history data including the post-interpolated statistical data extracted from the time series of the post-interpolated history data in the order of time indicated by the post-interpolated time data;
outputting the compression history data to a database;
Data compression method. In calculating the time series of the interpolated history data, time data and statistical data constituting the n-th history data in the time series of the history data are t ⁿ and v ⁿ , and the n-th time series in the time series of the interpolated history data is When the interpolated time data and the interpolated statistical data that constitute the interpolated history data are t ^n' and v ^n' , and the collection interval is C,
t ^n' = t ^n-1' + C...(1)
v ^n' , = (1-α) x v ^n-1' + α x v ⁿ , ... (2)
α=(t ^n' - t ^n-1' )/(t ⁿ - ^{t n-1'} ) ... (3)
Calculating the time series of the interpolated history data by the following calculation,
The data compression method according to claim 6. In the calculation of the compressed history data, the interpolated statistical data in the compressed history data is further compressed by performing an XOR operation on two pieces of the interpolated statistical data in floating point format.
The data compression method according to claim 7. further reading the compressed history data from the database, decompressing it, and outputting the interpolated history data;
The data compression method according to claim 6. to the computer,
A time series of historical data including time data and statistical data associated with the time data, and data indicating a predetermined collection interval are correlated with the post-interpolated time data and the post-interpolated time data by an interpolation operation. a process of calculating a time series of interpolated historical data including post-interpolated statistical data;
a process of calculating compressed history data including the post-interpolated statistical data extracted from the time series of the post-interpolated history data in the order of time indicated by the post-interpolated time data;
a process of outputting the compression history data to a database;
A program to do this.

Images