JP2019035995A - Two-dimensional data analysis apparatus, two-dimensional data analysis method, and two-dimensional data analysis program - Google Patents

Abstract

To provide a two-dimensional analysis apparatus, two-dimensional data analysis method and a two-dimensional analysis program for efficiently analyze two-dimensional data at high speed and efficiently.SOLUTION: An analysis apparatus 102 analyzes an analysis value by a moment which is a parameter representing a feature of an analysis value in a plurality of combinations calculated using a power of the analysis value based on a combination of two-dimensional data and analysis values that are values associated with two-dimensional data, and includes a data acquisition unit 210 for acquiring a plurality of combinations, a preprocessing data calculation unit 231 for calculating the power of the analysis value in the acquired combination and a cumulative sum of powers, an analysis request input unit 220 for receiving a select of at least one of combinations among the acquired combinations, and a moment calculation unit 232 for calculating a moment of a selected analysis value which is an analysis value in the selected combination based on the calculated cumulative sum.SELECTED DRAWING: Figure 2

Description

  The present invention relates to a two-dimensional data analysis apparatus, a two-dimensional data analysis method, and a two-dimensional data analysis program.

  One of the visualization tools in data analysis is a heat map. The heat map is created by, for example, associating two-dimensional data with an area composed of a plurality of rectangular sections and calculating the statistics of each data.

  Such data analysis is attracting attention as utilization of so-called big data in recent years. As a technology related to this, Japanese Patent Application Laid-Open No. H10-228561 generates a substitute image in advance for speeding up display of the heat map, and searches for and provides a corresponding substitute image when requested by the user. It is described that the generation of the substitute image is performed only within the range set by the substitute range setting means.

JP 2012-79037 A

  However, Patent Document 1 does not speed up the analysis in a region outside the region where the heat map substitute image exists, and does not assume data analysis by a method other than creation of the heat map. Therefore, in this document, there is a case where data processing cannot be performed efficiently and big data cannot be handled.

  The present invention has been made in view of such a background, and an object thereof is a two-dimensional data analysis apparatus, a two-dimensional data analysis method, and a two-dimensional data analysis for analyzing two-dimensional data at high speed and efficiently. To provide a program.

  One of the present invention for achieving the above object is calculated using a power value of the analysis value based on a combination of two-dimensional data and an analysis value that is a value associated with the two-dimensional data, An analysis apparatus for analyzing the analysis value based on a moment that is a parameter representing the characteristics of the analysis value in a plurality of the combinations, the data acquisition unit acquiring a plurality of the combinations, and the analysis in the acquired combination A pre-processing data calculation unit that calculates a power value of values and a cumulative sum related to the power value, an analysis request input unit that receives selection of at least one of the acquired combinations, and the selected Moment to calculate the moment of the selected analysis value, which is the analysis value in the combination, based on the calculated cumulative sum It includes a detecting section, a.

According to the present invention, two-dimensional data can be analyzed quickly and efficiently.
Problems, configurations, and effects other than those described above will be clarified by the following description of embodiments.

FIG. 1 is a diagram illustrating an example of a hardware configuration in the analysis system 100 according to the present embodiment. FIG. 2 is a diagram illustrating an example of functions provided in the analysis apparatus 102. FIG. 3 is a diagram illustrating an example of the management table 230. FIG. 4 is a diagram illustrating an example of the temporary DB 231. FIG. 5 is a flowchart for explaining an example of the two-dimensional data analysis process. FIG. 6 is a diagram illustrating an example of the data input screen. FIG. 7 is a diagram illustrating an example of the analysis data file. FIG. 8 is a flowchart illustrating an example of the preprocess data calculation process. FIG. 9 is a flowchart for explaining the details of the heat map data generation process. FIG. 10 is a diagram showing an example of an analysis screen 700 displayed in the heat map data generation process. FIG. 11 is a diagram illustrating an example of an analysis system 10 to which the analysis apparatus 102 according to the present embodiment is applied.

  Hereinafter, embodiments of the present invention will be described in detail with reference to the drawings. In the present specification, unless otherwise specified, addition / subtraction / multiplication / division operations are represented by “+”, “−”, “*”, and “/”, respectively. The power is represented by “^”, and the sum of data is represented by “Σ”.

  The analysis system of the present embodiment creates a two-dimensional diagram representing the distribution or characteristics of predetermined measurement values or analysis values (hereinafter referred to as analysis values) such as a so-called heat map.

  In this embodiment, a moment is used as a numerical parameter for representing the distribution or characteristics of the analysis value. The moment here is a parameter calculated using a power value of the analysis value, as will be described later.

<Hardware configuration>
FIG. 1 is a diagram illustrating an example of a hardware configuration in the analysis system 100 according to the present embodiment. As shown in the figure, the analysis system 100 includes an input device 101 composed of a keyboard, a mouse, a touch panel, etc., an analysis device 102 which is an information processing device (computer), and an output device 103 composed of a monitor (display). It is configured to include. The analysis apparatus 102 includes a processor 51 such as a CPU (Central Processing Unit), an R
A main storage device 52 such as an AM (Random Access Memory) and a ROM (Read Only Memory) and an auxiliary storage device 53 such as an HDD (Hard Disk Drive) and an SSD (Solid State Drive) are provided.

  In the example of FIG. 1, the input device 101, the analysis device 102, and the output device 103 are shown as separate devices, but these may be integrally configured as the analysis device 102. Therefore, in the present embodiment, it is assumed that the analysis device 102 is integrally configured including the input device 101 and the output device 103.

<Function>
Next, functions provided in the analysis apparatus 102 will be described.
The analysis device 102 calculates a power value of the analysis value based on a combination of the two-dimensional data corresponding to the coordinates in the heat map and the analysis value that is a value associated with the two-dimensional data (hereinafter referred to as analysis data). It is an analyzer for analyzing an analysis value by a moment that is a parameter that is calculated by using and represents a characteristic of an arbitrary analysis value in a plurality of analysis data.

  The functions of the analysis device 102 described below are executed by the hardware of the analysis device 102 or by the processor 51 of the analysis device 102 reading and executing each program stored in the main storage device 52 or the auxiliary storage device 53. Is realized. In addition, these programs are, for example, a secondary storage device, a nonvolatile semiconductor memory, a hard disk drive, a storage device such as an SSD, or a non-transitory data storage that can be read by a computer, such as an IC card, an SD card, a DVD Stored on media.

  FIG. 2 is a diagram illustrating an example of functions provided in the analysis apparatus 102. As shown in the figure, the analysis apparatus 102 includes a data acquisition unit 210, a preprocessing data calculation unit 231, a preprocessing data storage unit 213, an analysis request input unit 220, a moment calculation unit 232, and a result output unit 224.

  The data acquisition unit 210 acquires a plurality of the combinations.

  For example, the data acquisition unit 210 receives input of two-dimensional coordinate data and a combination of analysis values (analysis data) at the coordinates from an administrator or user of the analysis apparatus 102 (hereinafter referred to as a user).

  The preprocessing data calculation unit 231 calculates a power value of the analysis value in the combination acquired by the data acquisition unit 210 and a cumulative sum related to the power value.

  Specifically, the preprocessing data calculation unit 231 uses, as the data power of the analysis value in the combination (analysis data) acquired by the data acquisition unit 210, a power value by each exponent that is a predetermined maximum order or less, The acquisition unit 210 calculates each of the acquired plurality of analysis values.

  For example, when the degree of the power value is 2, the power value by each exponent is a power value and a power value.

  The analysis request input unit 220 accepts selection of at least one or more of the combinations acquired by the data acquisition unit 210.

  That is, the analysis request input unit 220 receives input of analysis data to be displayed as a heat map among the analysis data input in advance by the data acquisition unit 210. It should be noted that all analysis data input by the data acquisition unit 210 may be displayed as a heat map.

  The moment calculation unit 232 calculates the moment of the selected analysis value that is the analysis value in the combination selected by the analysis request input unit 220 based on the calculated cumulative sum.

  The moment is, for example, a power value based on a predetermined exponent, of the difference between the analysis value in the plurality of combinations and the average value of the analysis values in the plurality of combinations.

  That is, when arbitrary data is x, the average value of all data is u, and the moment order is n, the moment is Σ [(x−u) ^ n] (where Σ represents the sum of all data). When n is 1, the moment is defined as u itself.

The moment calculation unit 232 includes a cumulative sum reading unit 222. The cumulative sum reading unit 222 reads various data relating to the cumulative sum from the management table 230 or the temporary DB 231 described later.

  The moment calculation unit 232, when calculating the moment of the selection analysis value, the processing request decomposition unit 221 that preconfigures the moment of the selection analysis value as a linear sum of the cumulative sum of the selection analysis value, and the configuration And a cumulative sum calculator 223 for calculating a moment of the selected analysis value based on the calculated linear sum and the calculated cumulative sum.

  Specifically, when calculating the moment of the selected analysis value, the processing requirement decomposing unit 221 uses the moment of the selected analysis value as a cumulative sum related to the power value of each of the selected analysis values by the power exponent. Preconfigured as a linear sum.

  The cumulative sum calculation unit 223 calculates the moment of the selected analysis value by associating the calculated sum of the exponents with the linear sum constructed by the processing request decomposition unit 221.

For example, the analysis request input unit 220 decomposes the moment expression Σ [(xu) ^ 2] of the selected analysis value to obtain [1 * Σ (x ^ 2)]-[2 * (uΣx)] + Σ ( u ^ 2) = [1 * Σ (x ^ 2)]-(1 / N) * [(N * u) ^ 2]

  The result output unit 224 is a two-dimensional diagram showing a relationship between the two-dimensional data in the combination selected by the analysis request input unit 220 and a statistical value calculated by the calculated moment of the selected analysis value. Output.

  Specifically, the result output unit 224 displays the heat map that is the two-dimensional diagram showing the characteristics of the selected analysis value in a display format corresponding to the statistical value.

  The statistical value is, for example, a standard deviation when the moment is a second-order moment. Further, when the moment is a third-order moment, it is a skewness.

  The analysis apparatus 102 includes a management table 230 that is a database that manages information related to heat maps and cumulative sums, and a temporary DB 231 that is a database that stores detailed information on cumulative sums.

Here, the management table 230 will be described.
(Management table 230)
FIG. 3 is a diagram illustrating an example of the management table 230. The management table 230 includes a data name 612 and a data name 612 in which names (identification information) of databases (hereinafter referred to as analysis databases) of various analysis data (for example, land price distribution data or population distribution data in Japan) are stored. In the analysis database indicated by, the cumulative sum ID 613 for storing information for identifying the cumulative sum of analysis values (hereinafter referred to as cumulative sum ID), and the exponent (the power indicated by the cumulative sum ID 613) in the cumulative sum indicated by the cumulative sum ID 613 A database composed of at least one record having each item of index 611 in which the sum is an n-th cumulative sum (a value of n when the sum is an n-th value of the analysis value) is there.

Next, the temporary DB 231 will be described.
(Temporary DB 231)
FIG. 4 is a diagram illustrating an example of the temporary DB 231. As shown in the figure, the temporary DB 231 is a table that stores the details of the cumulative sum of analysis values, the cumulative sum ID 621 that stores the cumulative sum ID corresponding to the cumulative sum ID 613 of the management table 230, and the cumulative sum. This is a database composed of at least one record having each item of accumulated sum data 622 in which accumulated data of analysis values indicated by ID 621 is stored. In the cumulative sum data 622, for example, a cumulative sum related to each analysis value is stored as array data.

Next, processing performed by the analysis apparatus 102 will be described.
<Processing>
FIG. 5 is a flowchart for explaining an example of processing (hereinafter referred to as “two-dimensional data analysis processing”) performed by the analysis apparatus 102 to create a two-dimensional diagram such as a heat map from analysis data. This process is executed, for example, when the analysis apparatus 102 is activated.

  First, the data acquisition unit 210 of the analysis apparatus 102 displays a predetermined menu screen (not shown) (s300) and inputs analysis data from the user or analyzes analysis data based on the input analysis data. Selection of whether to perform (create heat map data) is accepted (s301).

  When the selection to input the analysis data is made (s301: data input), the data acquisition unit 210 receives the input of the analysis data (s304), and thereafter the process of s305 is performed. On the other hand, when the analysis data is selected to be analyzed (s301: data analysis), the process of s302 described later is performed.

(Data entry screen)
As an example of the process performed in s304, FIG. 6 is a diagram illustrating an example of a screen that accepts input of analysis data (hereinafter referred to as a data input screen). As shown in the figure, the data input screen 300 includes an analysis database and information for specifying the analysis data (for example, the name of the analysis database and the name of a file storing each analysis data (hereinafter referred to as an analysis data file)). Each field of an input file designation field 301, an upload button 302 for registering the analysis database designated in the file designation field 301, and a maximum order designation button 303 for designating the maximum degree N of moment used for analysis of analysis data Have The inputted information is stored in the management table 230.

Here, FIG. 7 is a diagram illustrating an example of the analysis data file. As shown in the figure, the analysis data file 400 is a CSV file (CSV: comma-separated values), for example, and stores two-dimensional data and a combination of analysis values corresponding to the two-dimensional data. In the example shown in the figure, each column 401 delimited by a comma is a first element of two-dimensional data, each row 402 delimited by a line feed is a second element of two-dimensional data, and each value 403 in the columns 401 and 402 is It becomes an analysis value. For example, the column 401 corresponds to the x coordinate in the two-dimensional coordinate system (x, y), and the row 402 corresponds to the y coordinate in the two-dimensional coordinate system (x, y).

  Next, as shown in FIG. 5, the preprocess data calculation unit 231 performs a process of calculating a cumulative sum related to the analysis value input in s304 (hereinafter referred to as a preprocess data calculation process) (s305). Details of this processing will be described later. Thereafter, the process of s300 is performed.

  Note that the processes of s304 and s305 are repeated each time there is a need to modify, add, delete, etc., for example, the input analysis data (two-dimensional data).

  On the other hand, in s302, the moment calculation unit 232 generates data (for example, moment) necessary for creating the heat map based on the cumulative sum calculated by the preprocessing data calculation process (hereinafter referred to as heat map data generation process). I do. Details of this processing will be described later.

Then, the result output unit 224 generates a heat map based on the data calculated in the heat map data generation process and displays the heat map on a predetermined screen (s303). By viewing this screen, the user grasps the distribution or characteristics of the analysis data. Thereafter, the process of s300 is performed.

Next, details of the preprocess data calculation process and the heat map data generation process will be described.
<Preprocessing data calculation process>
FIG. 8 is a flowchart illustrating an example of the preprocess data calculation process. As shown in the figure, first, the preprocessing data calculation unit 231 initializes a value of a predetermined variable n (hereinafter referred to as a counter n) indicating the number of processing repetitions to 0 (s401), and stores that fact.

  The power calculation unit 211 determines whether or not the current value of the counter n is less than or equal to the maximum moment order N (s402), and if the current value of the counter n is less than or equal to the maximum moment order N ( (s402: YES), the value of the counter n is incremented by 1 (s403), and then the process of s404 is performed. On the other hand, when the current value of the counter n exceeds the maximum moment order N (s402: NO), this process ends (s408).

  In s404, the power calculation unit 211 obtains a power value related to each analysis value input in s304.

For example, when the input analysis data is [5, 4, 1], which is a set of analysis values of two-dimensional data of 1 row and 3 columns, the power calculation unit 211 calculates the first power value [5, 4, 1], the square value of each analysis value [25,
16, 1], the cube of each analysis value [125, 64, 1], ..., and the N-th power value [5 ^ N, 4 ^ N, 1 ^ N] of each analysis value .

  Then, the cumulative sum calculation unit 212 calculates the nth power sum of each analysis value using the power value calculated in s403 (s405).

  Specifically, for example, the cumulative sum calculation unit 212, in s404, the nth power value x [i] [j] corresponding to the two-dimensional matrix (H rows and W columns) (i is 0 or more (H-1) When the following integer, j is an integer greater than or equal to 0 and not greater than (W-1), the cumulative sum calculation unit 212 uses this to calculate the nth power sum in (H + 1) rows (W + 1) columns y [k] [l] (k is an integer from 0 to H, and l is an integer from 0 to W) is calculated as follows.

  First, the cumulative sum calculation unit 212 sets y [0] [l] = 0 and y [k] [0] = 0 for all k and l.

Next, the cumulative sum calculation unit 212, when i = 0 is fixed,
y [i + 1] [j + 1] = x [i] [j] + y [i + 1] [j] (1)
When j is set to 0, 1 is set,..., W-1 is performed in order, y [1] [0], y [1] [1], ..., y [1] [W-1] is calculated respectively.

Similarly, the cumulative sum calculation unit 212
y [i + 1] [l] = y [i] [l] + y [i + 1] [l] (2)
When l is set to 0, 1 is set, ..., W is set in order, y [1] [0], y [1] [1], ..., y [ 1] Calculate [W] respectively.

  Such calculation of the array y according to the equations (1) and (2) is sequentially performed for the case where i = 1, 2, 3,. As a result, the finally calculated value y [k] [l] is the n-th cumulative sum corresponding to x [i] [j].

When the calculation of the n-th cumulative sum is completed, the preprocessing data storage unit 213 manages each analysis data input in s304, the names of these analysis data, and the current exponent (value of the counter n). It memorize | stores in the table 230 (s406).

  Specifically, the preprocessing data storage unit 213 generates a new record in which a predetermined cumulative sum ID is assigned to the cumulative sum ID 613 of the management table 230, and sets the current value of the counter n to the index 611 of the generated record. The name of the analysis data (set) is stored in the data name 612.

  Further, the preprocessing data storage unit 213 stores the information on the n-th power sum calculated in s404 in the temporary DB 213 (s407).

Specifically, the preprocessing data storage unit 213 generates a record in which the cumulative sum ID is set to the cumulative sum ID 621 in the temporary DB 231, and adds each nth power calculated in s 405 to the cumulative sum data 622 of the generated record. Stores the cumulative sum value (array y). Thereafter, the process of s402 is performed.

  In this way, the power calculation unit 211 obtains a power value of each analysis value by each power index of 1 or more and the maximum moment order N or less.

Next, heat map data generation processing will be described.
<Heat map data generation process>
FIG. 9 is a flowchart for explaining the details of the heat map data generation process.

  First, the processing request decomposing unit 221 displays an analysis screen to be described below, thereby receiving input of information related to data necessary for creating a heat map (hereinafter referred to as an analysis request) from the user (s500).

<Analysis screen>
Here, FIG. 10 is a diagram illustrating an example of the analysis screen 700 displayed in the heat map data generation process. As shown in the figure, this analysis screen 700 generates a heat map in a database selection field 705 that accepts input of an analysis database (hereinafter referred to as a selection database) in which analysis data to be analyzed is stored. A selection range input field 706 for receiving an input of a range of dimension data (hereinafter referred to as a selection range), and an analysis type input field for receiving an input of an analysis type (specifically, a type of moment, hereinafter referred to as an analysis type) It has 707 columns.

  The analysis screen 700 also includes a heat map display field 702 that displays the generated heat map. Details of the heat map display field 702 will be described later.

  The user inputs an analysis request using such an analysis screen 700.

  Next, as shown in s501 in FIG. 9, the processing request decomposing unit 221 configures each moment in the selection range in the form of a linear sum of power sums. Specifically, the following processing is performed.

  First, the processing request decomposing unit 221 stores the maximum degree of the moment specified by the analysis type (hereinafter referred to as the maximum order. In the present embodiment, the secondary order).

  Further, the cumulative sum reading unit 222 acquires all analysis values (selection analysis values) in the selection range from the management table 230.

Then, the cumulative sum calculation unit 223 calculates the moment of the analysis value (selection analysis value) in the selection range (secondary moment in this embodiment) as follows. That is, the second moment according to the analysis value in the selected range is
Σ (x _i -u) ^ 2 (x is each analysis value i in the selection range, u is the average value of all analysis values in the selection range) (3)
It can be expressed by the formula relating to the sum of the second power sums.

Here, the processing request decomposing unit 221 configures the expression (3) in the form of a linear sum of the power sum of each exponent. For example, if the number of analysis values in the selected range is N, equation (3) is
Σ (xu) ^ 2 = 1 * Σ (x ^ 2) -2 * (uΣx) + Σ (u ^ 2) = 1 * Σ (x ^ 2) -1 / N * (N * u) ^ 2 ・(4)
Can be configured.

  That is, the equation (4) can be configured as a linear sum of terms obtained by multiplying a power sum by a power exponent such as a square sum Σ (x ^ 2) and a power sum N * u = Σx by a predetermined coefficient.

  In this way, by configuring the moment of the analysis value as a linear sum of each power sum, the moment of each analysis value in the selection range can be calculated quickly, regardless of the selection range, as described later. And can be performed efficiently.

  Subsequently, the cumulative sum calculation unit 223 initializes a counter n (which is different from n used in the preprocessing data calculation process), which is a variable, and a variable ret that stores a moment calculation result, respectively, with 0 (s502). . Then, the cumulative sum calculation unit 223 calculates the n-th cumulative sum by repeatedly executing the processing from s503 to s507 described below based on the counter n.

  That is, first in s503, the cumulative sum calculation unit 223 confirms whether or not the current value of the counter n is equal to or less than the maximum order. If the current value of the counter n is less than or equal to the maximum order, the process proceeds to s504 (s503: YES), and if the current value of the counter n exceeds the maximum order (s503: NO), the process proceeds to s508. .

  In s504, the cumulative sum calculation unit 223 increments the value of the counter n by 1.

  Then, the cumulative sum reading unit 222 acquires all n-th cumulative sums of analysis values in the selected database from the temporary DB 231 (s505).

  Specifically, for example, the cumulative sum reading unit 222 acquires the cumulative sum ID 613 of the record of the management table 230 in which the selected database is recorded in the data name 612 and the value of the counter n is recorded in the index 611. All the contents of the cumulative sum data 622 of the record of the temporary DB 231 in which the cumulative sum ID same as the cumulative sum ID indicated by the cumulative sum ID 613 is stored in the cumulative sum ID 621 are acquired.

  Then, the cumulative sum calculation unit 223 calculates all the sums (specifically, the nth power sum) of the selected analysis values based on the nth power cumulative sum acquired in s505 (s506).

The n-th power sum is calculated as follows, for example. When the upper left coordinate in the selection range is (r1, c1) and the lower right coordinate in the selection range is (r2, c2),
y [r2 + 1] [c2 + 1] -y [r2 + 1] [c1] -y [r1] [c2 + 1] + y [r1] [c1] (5)
It is expressed.

Here, since each array y in equation (5) is the n-th power cumulative sum acquired in s505, the cumulative sum calculator 223 can easily obtain the n-th power sum in the selection range using formula (5). .

That is, in order to obtain the sum of n powers in a predetermined selection range, conventionally, for example, (r2-r1 + 1) * (c2-c1 + 1) times corresponding to the number of analysis values in the selection range. Calculation was necessary. For this reason, the amount of calculation is proportional to the number of analysis values (selection range), which requires a lot of calculation time.

However, in the present embodiment, the cumulative sum calculation unit 223 selects only three operations related to the array y by substituting into the equation (5) using the cumulative sum (n-th power cumulative sum) calculated in s505. Range n
Multiplicative sum can be calculated. For example, when the upper left coordinate of the selection range is (301, 501) and the lower right coordinate is (400, 600), it is normally necessary to calculate 10,000 times, but the cumulative sum is calculated as in this embodiment. If used, the calculation can be completed in three times. That is, the number of moment calculations can be made constant irrespective of the data size (the size of the selection range).

  Next, the cumulative sum calculation unit 223 multiplies the coefficient (see Expression (4)) corresponding to the n-th power calculated in s506, and adds the multiplication result (product) to the value of the current variable ret (s507). ). Thereafter, the process of s503 is performed again.

  On the other hand, when the current value of the counter n exceeds the maximum order in s503 (s503: NO), the result output unit 224 ends the calculation of the n-th power and displays the heat map related to the selected range. The statistical value is generated (s508).

That is, the current variable ret stores the second moment value (Σ [(xu) ^ 2]) of the selection analysis value in the selection range. Therefore, for example, the result output unit 224 calculates the standard deviation of the selection analysis value in the selection range based on the value of the variable ret.

  Then, the result output unit 224 identifies a display method on the heat map (for example, a display color set in advance by the range of the standard deviation value) that is associated with the standard deviation value calculated above. Is the display format in the selection range.

  Thus, the heat map data generation process ends (s509). Note that the heat map data generation processing described so far is performed for all of the selection ranges when a plurality of selection ranges are designated in s500.

  Here, the example of the heat map produced | generated by the above process is demonstrated using the said analysis screen 700. FIG.

<Display of heat map in heat map display field 702>
The heat map shown in the heat map display field 702 in FIG. 10 is a heat map represented by a predetermined section 701 in which a delay time of a predetermined region is divided for each date and each time zone for arrival at a distributor. Yes, one axis 703 represents the date, and the other axis 704 represents the time zone.

  Each section 701 has a display corresponding to the selection range, that is, a display format display (for example, color coding) according to the statistical value of the selection range.

In the example of FIG. 10, the selection range input field 706 shows the entire range from the upper left coordinate to the lower right coordinate of the heat map, that is, the range of all sections 701. As described above, in specifying a range in the selection range input field 706, the analysis request input unit 220 uses a plurality of ranges selected in the selection range input field 706 at a predetermined resolution (for example, preset by the user). It may be divided into squares (sections 701), and each of the divided sections 701 may be selected. By doing in this way, the user can specify easily the part which wants to display a heat map with desired resolution.

  The user refers to the heat map display field 702 and compares it with a conventional heat map (a heat map in which an average delay time is displayed for each area), so that an area where the delay is large and the standard deviation is small is constant. It is possible to determine that a region where a delay occurs and a region where the delay is large and the standard deviation is large is a region where there are many sudden delays.

  That is, the conventional heat map alone can only grasp the average delay time for each region, and cannot grasp the specific cause of the delay in each region. However, by using the heat map of the present embodiment, it is possible to grasp the characteristics of the delay and analyze the specific cause of the delay. For example, the heat map display field 702 allows the user to determine that many constant delays have occurred, and in this case, it can be estimated that there may be a problem with the shipping system itself. In addition, since the user can also determine from the heat map display field 702 that many sudden delays have occurred, in that case, there is a possibility that the delay of arrival is basically due to traffic conditions or the like. It can be estimated that there is. In this way, the user can execute an appropriate countermeasure against delay by referring to the heat map display field 702.

<Example of analysis system>
In addition, if the analysis apparatus 102 of this embodiment is used, it is also possible to construct | assemble the analysis system which produced | generated the suitable heat map efficiently according to the request | requirement from each user.

FIG. 11 is a diagram illustrating an example of an analysis system 10 to which the analysis apparatus 102 according to the present embodiment is applied. As shown in the figure, the analysis system 10 can communicate with the analysis apparatus 102 of this embodiment via a communication network 108 such as a LAN (Local Area Network), a WAN (Wide Area Network), the Internet, a dedicated line, or the like. And one or a plurality of user terminals 105 (information processing devices) connected to each other. The analysis apparatus 102 generates a heat map corresponding to the analysis request in response to the analysis request transmitted from each user terminal 105, and transmits the generated heat map to each user terminal 105. Each user terminal 105 displays the heat map received from the analysis apparatus 102. At this time, each user can also select a heat map to be used using the management table 230.

  As described above, according to the analysis apparatus 102 of the present embodiment, the power value of the analysis value in a plurality of combinations of two-dimensional data and analysis values (analysis data) and the cumulative sum related to the power value are calculated in advance. Calculate (pre-processing), accept selection (selection range) of at least one combination from the acquired combinations, and select the moment of the selected analysis value in the selected combination to the cumulative sum calculated in the pre-processing. Therefore, the moment of the selected analysis value can be obtained quickly and efficiently based on the cumulative sum calculated in advance in the preprocessing. In other words, the moment is a parameter that is calculated using the power value of the analysis value, so if the cumulative sum of the analysis value is obtained in advance in the preprocessing, the power value can be calculated at high speed by using this cumulative sum. The power value of the selected analysis value can be easily obtained. Thereby, a statistical value (standard deviation or skewness) of an arbitrary selection range can be obtained quickly and efficiently.

For example, in a heat map analysis, when a heat map is created and displayed many times while changing the analysis range and resolution, calculation of moment according to the resolution is performed each time the resolution is changed. Since it had to be repeated, the amount of calculation was very large and the processing took time. Therefore, the data amount (number of analysis values) itself has to be reduced to a size that can withstand the display of the heat map. However, since the analysis apparatus 102 according to the present embodiment obtains the cumulative sum of analysis values in the preprocessing, the selected range can be used even when the amount of data is large (for example, big data). The moment can be calculated at high speed, and the heat map can be displayed quickly.

  In addition, since there are various types of moments in the present embodiment, appropriate processing according to the purpose of analysis can be performed quickly. For example, the analysis can be performed from various viewpoints such as the standard deviation (dispersion) of the analysis value for the second moment, and the skewness for the third moment. According to the analysis apparatus 102 of the present embodiment, It is possible to quickly respond to such various types of analysis requests.

  Furthermore, the analysis apparatus 102 of the present embodiment configures the moment of the selected analysis value as a linear sum of the cumulative sum of the selected analysis values, and calculates the moment of the selected analysis value based on this linear sum and the cumulative sum calculated above. Therefore (Equation (4)), the moment can be calculated at high speed with a simple calculation.

  For example, when the moment is expressed as a power value by a predetermined power index of the difference between the analysis value and the average value of all the analysis values, the analysis apparatus 102 of the present embodiment uses the power value as a predetermined maximum order or less. The moment of the selected analysis value is calculated by calculating the power value of each of the exponents, and the moment of the selected analysis value corresponding to the cumulative sum of the power of the selected analysis value according to the exponent. By simplifying the moment calculation in this way, even when the analysis data is a large amount of data such as so-called big data, data analysis can be performed at high speed.

  In addition, the analysis apparatus 102 according to the present embodiment outputs a two-dimensional diagram (heat map) indicating the relationship between the two-dimensional data and the statistical value calculated by the moment of the selected analysis value. Based on the output result, the distribution or characteristics of each analysis value can be easily grasped and used for analysis.

  For example, the analysis apparatus 102 according to the present embodiment displays a heat map indicating the characteristics of the selected analysis value in a display format corresponding to the statistical value, so that the user can change each analysis value based on the difference in the display format. Distribution or characteristics can be easily grasped.

  The above description of the embodiment is for facilitating the understanding of the present invention, and does not limit the present invention. The present invention can be changed and improved without departing from the gist thereof, and the present invention includes equivalents thereof.

  For example, each of the above-described configurations, functions, processing units, processing means, and the like may be realized by hardware by designing a part or all of them with, for example, an integrated circuit.

  In this embodiment, an example of a heat map has been described as a two-dimensional diagram represented by two-dimensional data and analysis values. However, the two-dimensional diagram is not limited to this, and the distribution or characteristics of analysis values can be expressed two-dimensionally. Any diagram can be applied to other types of diagrams.

  In this embodiment, a database (table) is used as a data storage format, but other storage formats may be used.

DESCRIPTION OF SYMBOLS 100 Analysis system, 102 Analysis apparatus, 210 Data acquisition part, 231 Preprocessing data calculation part, 220 Analysis request input part, 232 Moment calculation part

Claims (12)

A parameter that is calculated using a power value of the analysis value based on a combination of two-dimensional data and an analysis value that is a value associated with the two-dimensional data, and represents a characteristic of the analysis value in a plurality of the combinations An analysis device comprising a processor and a memory for analyzing the analysis value by a certain moment,
A data acquisition unit for acquiring a plurality of the combinations;
A pre-processing data calculation unit for calculating a power value of the analysis value in the acquired combination and a cumulative sum related to the power value;
An analysis request input unit that accepts selection of at least one of the acquired combinations from the combination;
A moment calculator that calculates the moment of the selected analysis value, which is the analysis value in the selected combination, based on the calculated cumulative sum;
A two-dimensional data analysis apparatus. The moment calculator is
When calculating the moment of the selection analysis value, a processing request decomposition unit that preconfigures the moment of the selection analysis value as a linear sum of the cumulative sum of the selection analysis values;
The two-dimensional data analysis apparatus according to claim 1, further comprising: a cumulative sum calculation unit that calculates a moment of the selected analysis value based on the configured linear sum and the calculated cumulative sum. The moment is a power value by a predetermined exponent of the difference between the analysis value in the plurality of combinations and the average value of the analysis values in the plurality of combinations,
The preprocessing data calculation unit calculates, for each of the acquired plurality of analysis values, a power value by each exponent that is a predetermined maximum order or less as a power value of the analysis value in the acquired combination,
When calculating the moment of the selected analysis value, the processing request decomposing unit preconfigures the moment of the selected analysis value as a linear sum of a cumulative sum related to a power value of each of the selected analysis values by the power exponent. ,
The cumulative sum calculation unit calculates the moment of the selected analysis value by associating the calculated sum of the exponents with the calculated linear sum.
The two-dimensional data analysis apparatus according to claim 2.   The apparatus according to claim 1, further comprising a result output unit that outputs a two-dimensional diagram showing a relationship between the two-dimensional data in the selected combination and a statistical value calculated by the calculated moment of the selected analysis value. The two-dimensional data analysis apparatus described.   The two-dimensional data analysis apparatus according to claim 4, wherein the result output unit displays a heat map that is the two-dimensional diagram showing the characteristics of the selected analysis value in a display format corresponding to the statistical value. A result output unit that outputs a two-dimensional diagram showing a relationship between the two-dimensional data in the selected combination and a statistical value calculated by the calculated moment of the selected analysis value;
The moment is a power value by a predetermined exponent of the difference between the analysis value in the plurality of combinations and the average value of the analysis values in the plurality of combinations,
The preprocessing data calculation unit calculates, for each of the acquired plurality of analysis values, a power value by each exponent that is a predetermined maximum order or less as a power value of the analysis value in the acquired combination,
The moment calculator is
When calculating the moment of the selection analysis value, a processing request decomposition unit that preconfigures the moment of the selection analysis value as a linear sum of the cumulative sum of the selection analysis values;
A cumulative sum calculator that calculates the moment of the selected analysis value based on the configured linear sum and the calculated cumulative sum;
When calculating the moment of the selected analysis value, the processing request decomposing unit preconfigures the moment of the selected analysis value as a linear sum of a cumulative sum related to a power value of each of the selected analysis values by the power exponent. ,
The cumulative sum calculation unit calculates a moment of the selected analysis value by associating the calculated sum of the exponents with the calculated linear sum,
The result output unit displays a heat map that is the two-dimensional diagram showing the characteristics of the selected analysis value in a display format corresponding to the statistical value.
The two-dimensional data analysis apparatus according to claim 1. A parameter that is calculated using a power value of the analysis value based on a combination of two-dimensional data and an analysis value that is a value associated with the two-dimensional data, and represents a characteristic of the analysis value in a plurality of the combinations An analysis method for analyzing the analysis value by a certain moment,
An information processing apparatus comprising a processor and a memory
A data acquisition process for acquiring a plurality of the combinations;
A pre-processing data calculation process for calculating a power value of the analysis value in the acquired combination and a cumulative sum related to the power value;
An analysis request input process for accepting selection of at least one of the acquired combinations;
A moment calculation process for calculating the moment of the selected analysis value, which is the analysis value in the selected combination, based on the calculated cumulative sum;
A two-dimensional data analysis method. The moment calculation process includes:
When calculating the moment of the selection analysis value, processing requirement decomposition processing that pre-configures the moment of the selection analysis value as a linear sum of the cumulative sum of the selection analysis values;
The two-dimensional data analysis method according to claim 7, wherein a cumulative sum calculation process of calculating a moment of the selected analysis value based on the configured linear sum and the calculated cumulative sum is performed. The information processing apparatus includes:
A result output process for outputting a two-dimensional diagram showing a relationship between the two-dimensional data in the selected combination and a statistical value calculated by the calculated moment of the selected analysis value is further executed. 8. The two-dimensional data analysis method according to 7. A parameter that is calculated using a power value of the analysis value based on a combination of two-dimensional data and an analysis value that is a value associated with the two-dimensional data, and represents a characteristic of the analysis value in a plurality of the combinations An analysis program for analyzing the analysis value by a certain moment,
In an information processing apparatus including a processor and a memory,
A data acquisition process for acquiring a plurality of the combinations;
A pre-processing data calculation process for calculating a power value of the analysis value in the acquired combination and a cumulative sum related to the power value;
An analysis request input process for accepting selection of at least one of the acquired combinations;
A moment calculation process for calculating the moment of the selected analysis value, which is the analysis value in the selected combination, based on the calculated cumulative sum;
Is a two-dimensional data analysis program. The moment calculation process includes:
When calculating the moment of the selection analysis value, processing requirement decomposition processing that pre-configures the moment of the selection analysis value as a linear sum of the cumulative sum of the selection analysis values;
The two-dimensional data analysis program according to claim 10, wherein the two-dimensional data analysis program is configured to execute a cumulative sum calculation process for calculating a moment of the selected analysis value based on the configured linear sum and the calculated cumulative sum. In the information processing apparatus,
A result output process for outputting a two-dimensional diagram showing a relationship between the two-dimensional data in the selected combination and a statistical value calculated by the calculated moment of the selected analysis value is further executed. The two-dimensional data analysis program according to 10.

Images