CN107193862A - A kind of variance optimization histogram construction method and device based on Spark Streaming - Google Patents
A kind of variance optimization histogram construction method and device based on Spark Streaming Download PDFInfo
- Publication number
- CN107193862A CN107193862A CN201710212747.7A CN201710212747A CN107193862A CN 107193862 A CN107193862 A CN 107193862A CN 201710212747 A CN201710212747 A CN 201710212747A CN 107193862 A CN107193862 A CN 107193862A
- Authority
- CN
- China
- Prior art keywords
- data
- variance
- bucket
- window
- sampling
- Prior art date
- Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
- Pending
Links
- 238000005457 optimization Methods 0.000 title claims abstract description 93
- 238000010276 construction Methods 0.000 title claims abstract description 34
- 238000005070 sampling Methods 0.000 claims abstract description 123
- 238000000034 method Methods 0.000 claims abstract description 25
- 241001269238 Data Species 0.000 claims abstract description 10
- 238000013500 data storage Methods 0.000 claims description 12
- 238000009825 accumulation Methods 0.000 claims description 6
- 230000008859 change Effects 0.000 claims description 2
- 238000005516 engineering process Methods 0.000 description 9
- 238000004590 computer program Methods 0.000 description 7
- 238000010586 diagram Methods 0.000 description 6
- 238000009826 distribution Methods 0.000 description 5
- 238000012545 processing Methods 0.000 description 5
- 230000008901 benefit Effects 0.000 description 3
- 230000006870 function Effects 0.000 description 3
- 238000003860 storage Methods 0.000 description 3
- BYACHAOCSIPLCM-UHFFFAOYSA-N 2-[2-[bis(2-hydroxyethyl)amino]ethyl-(2-hydroxyethyl)amino]ethanol Chemical group OCCN(CCO)CCN(CCO)CCO BYACHAOCSIPLCM-UHFFFAOYSA-N 0.000 description 2
- 238000004458 analytical method Methods 0.000 description 2
- 238000004422 calculation algorithm Methods 0.000 description 2
- 238000013480 data collection Methods 0.000 description 2
- 239000002360 explosive Substances 0.000 description 2
- 230000006872 improvement Effects 0.000 description 2
- 238000012423 maintenance Methods 0.000 description 2
- 238000006116 polymerization reaction Methods 0.000 description 2
- 238000006467 substitution reaction Methods 0.000 description 2
- 230000007306 turnover Effects 0.000 description 2
- 241000282326 Felis catus Species 0.000 description 1
- 230000032683 aging Effects 0.000 description 1
- 238000004364 calculation method Methods 0.000 description 1
- 238000006243 chemical reaction Methods 0.000 description 1
- 238000004891 communication Methods 0.000 description 1
- 238000007405 data analysis Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000018109 developmental process Effects 0.000 description 1
- 230000000694 effects Effects 0.000 description 1
- 238000004519 manufacturing process Methods 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 238000012544 monitoring process Methods 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 230000002688 persistence Effects 0.000 description 1
- 230000008569 process Effects 0.000 description 1
- 238000011160 research Methods 0.000 description 1
- 238000007619 statistical method Methods 0.000 description 1
- 238000012360 testing method Methods 0.000 description 1
Classifications
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06F—ELECTRIC DIGITAL DATA PROCESSING
- G06F16/00—Information retrieval; Database structures therefor; File system structures therefor
- G06F16/90—Details of database functions independent of the retrieved data types
- G06F16/901—Indexing; Data structures therefor; Storage structures
- G06F16/9024—Graphs; Linked lists
Landscapes
- Engineering & Computer Science (AREA)
- Databases & Information Systems (AREA)
- Theoretical Computer Science (AREA)
- Software Systems (AREA)
- Data Mining & Analysis (AREA)
- Physics & Mathematics (AREA)
- General Engineering & Computer Science (AREA)
- General Physics & Mathematics (AREA)
- Complex Calculations (AREA)
Abstract
The present invention relates to a kind of variance optimization histogram construction method and device based on Spark Streaming, the method includes:On-line sampling is carried out using Spark Streaming stream datas;According to on-line sampling data, dynamic construction variance optimization histogram;The variance optimization histogram is updated using the new data dynamic that adds, and data dynamic construction variance optimization histogram again is added according to new;The technical scheme that the present invention is provided, can limited in memory headroom by the high-precision approximating variances optimization histogram of single pass data structure.
Description
Technical field
The present invention relates to big data calculating field, and in particular to a kind of variance optimization based on Spark Streaming is straight
Square figure construction method and device.
Background technology
With developing rapidly for internet, Internet of Things, cloud computing and the communication technology, the explosive growth of data is many
Industry brings brand-new opportunity to develop and challenge, for the emphasis being efficiently calculated as to study at present of mass data.Greatly
The computation schema of data can be divided into batch calculate and streaming computing both of which, due to flow data have continuation, it is ageing,
The characteristic such as sudden and data distribution is unknown, therefore calculated than the batch of off-line data, the online treatment technology of flow data
It is not perfect, and the demand field of the application such as efficient calculating and statistical analysis with the development stream data of technology is increasingly
It is high.It is just past 2016 that " the double 11 " activities of day cat, transaction peak value per second is up to 17.5 ten thousand, branch per second in first 30 minutes
Pay peak value 120,000,1 hour 57 second period turnover just more than 2013 double 11 whole days turnover;China Mobile pair
The user of 2014 has found that the talk times of average minute clock are up to 8,000,000 times, and the flow of mobile data is using data statistics
33GB per second, with 4G comprehensive covering and developing rapidly for 5G technologies, this numeral is also being skyrocketed through.Therefore, fluxion
According to treatment technology the numerous areas such as the real-time high-efficiency of network monitor, security monitoring and user data analysis are proposed more
High requirement.
One of method that flow data is commonly used in calculating is to build outline data structure, and the flow data that magnanimity is reached at a high speed is deposited
Storage is in specific outline data structure, come the outline data information for supporting quickly to obtain, statistic analysis result and high-precision
Approximate query.Variance optimization histogram therein is a kind of outline data structure of application field widely.Variance optimization is straight
Square figure is a kind of special histogram technology, the characteristics of variance sum with data in each bucket is minimum.Variance optimizes histogram
The profile of expression large data sets that not only can intuitively, succinctly, due to the characteristic that its variance is optimal, for any area on data set
Between aggregate query also there is very high precision.Therefore, variance optimization histogram in flow data high efficiency range aggregate query and
The fields such as data base querying are in occupation of very important status.
Traditional algorithm based on Dynamic Programming needs multiple ergodic data collection, therefore time and spatial complexity is larger;
Also one kind can build variance to arbitrary data collection within the sublinear time and optimize histogrammic method, but method can only be directed to from
Line number evidence, it is impossible to meet the histogrammic structure of variance optimization under flow data environment;It is limited for memory headroom under flow data environment
Premise, academia it is also proposed the histogram that a kind of utilization sample data builds variance optimization at present, but before the structure of the method
It is to know data distribution in advance to carry, can be to the flow data stochastical sampling that continuously reaches according to the Data distribution information of acquisition;
In addition, there is a kind of approximating variances optimization histogram method of dynamic adjustment at present, correspondence will be inserted per the element newly arrived
Bucket in, by the division to bucket and merge the variance sum near-optimization for make it that histogram is overall, the advantage of this method is significantly
Reducing structure variance and optimizing histogrammic time complexity, but have the disadvantage to need to preserve all initial data can just treat point
Split bucket and bucket to be combined carries out the calculating of variance, therefore the method is unfavorable for the dynamic structure in the case where limiting the streaming big data environment in space
Build variance optimization histogram;At present, under distributed computing framework, it is proposed that one kind is using MapReduce Computational frames to general
Data in rate database build approximating variances and optimize histogrammic method, but this method can only be counted for off-line data
Calculate, it is impossible to which rapid build is carried out to online flow data, and currently a popular flow data calculating platform does not also provide calculating side
The histogrammic method of difference optimization.
Accordingly, it is desirable to provide one kind can efficiently build variance under distributed stream data environment optimizes histogrammic method
To meet widespread need of the variance optimization histogram under flow data environment.
The content of the invention
The present invention provides a kind of variance optimization histogram construction method and device based on Spark Streaming, its mesh
Be that can build high-precision approximating variances optimization histogram in memory headroom by single pass data limiting.
The purpose of the present invention is realized using following technical proposals:
A kind of variance optimization histogram construction method based on Spark Streaming, it is theed improvement is that, including:
On-line sampling is carried out using Spark Streaming stream datas;
According to on-line sampling data, dynamic construction variance optimization histogram;
The variance optimization histogram is dynamically updated using the new data that add, and data dynamic construction again is added according to new
Variance optimizes histogram.
It is preferred that, it is described to carry out on-line sampling using Spark Streaming stream datas, including:
RDD time interval is set, is converted to flow data according to time interval using Spark Streaming
DStream structures, and window () operating parameter of RDD in DStream structures is called, the RDD in DStream structures is gathered
Window is combined into, then on-line sampling is carried out to the data in each window;
Wherein, the time interval of window is the positive integer times of RDD time interval.
Further, the data in each window carry out on-line sampling, including:
At the beginning of the sample data of the sample space of window is set as into K, and will sample threshold value T, variable sum and variable count
Beginning turns to 0, and greatest member number is initialized for K small top pile structure;
Data in the sample space of data storage in window to window, and in window are updated into the sampling threshold
Value T;
Wherein, when the sample data of the sample space of window is K+1, selected from the sample data of the sample space of window
Select the sample data ω for meeting constraintsiAnd ωj, and by sample data ωiAnd ωjThe small data accumulation of middle data value is to number
According to being worth in big data, the sample data ω is deletediAnd ωjThe small data of middle data value.
Further, the constraints includes:Sample data ωiWith ωjSum is less than the sampling threshold value T, and sample
Data ωiAnd ωjSampling cost it is minimum in the sample space of window;
Wherein, sample data ω is determined as the following formulaiAnd ωjSampling cost J:
In above formula, ωiFor i-th of sample data, ωjFor j-th of sample data, i<j;
Further, the data in window update the sampling threshold value T, including:
Judge whether the data value in window is less than sampling threshold value T;
If so, then the data value is added on variable sum, 1 is increased certainly with variations per hour count;
If it is not, then by the data storage into the small top pile structure;
Wherein, when the data amount check in small top pile structure reaches K, or the heap of small top pile structure serves as a fill-in the value of evidence and is less than and works as
During preceding sampling threshold value T, then the value that heap serves as a fill-in evidence is added on variable sum, with variations per hour count from increasing 1, and by the heap top
Data are deleted, and the value for the threshold value T that samples is updated into sum/count.
It is preferred that, the data according to on-line sampling, dynamic construction variance optimization histogram, including:
The maximum sample number that the sample space of set memory is allowed is KmaxAnd according to time interval setting variance optimization
The number of histogrammic bucket is B;
With the K first reached in the sample space of internal memorymaxThe data of individual online acquisition build the wide Nogata that bucket number is B
Figure, wherein, it is right to time interval institute where it to add the data of online acquisition according to the acquisition time of the data of online acquisition
The variance answered optimizes in histogrammic bucket, and determines the variance of each bucket using the data value in each bucket, builds histogram.
Further, when the data of the sample space of internal memory are KmaxAt+1, according to the sampling time of the data newly added
It is added into the variance to where it corresponding to time interval to optimize in histogrammic bucket, while updating the variance of this barrel, and selects
The bucket of variance maximum in current variance optimization histogram is selected as bucket to be divided, the minimum bucket of variance and its adjacent variance are smaller
A bucket as two buckets to be combined, if the variance of total data in two buckets to be combined is less than data in bucket to be divided
Variance, then perform split degree operation, conversely, not performing any operation then;
When the data of the sample space of internal memory are Kmax+ 1, and the data newly added sampling time not variance optimize
In the time zone of histogrammic bucket, then add it in the corresponding bucket of its sampling time nearest time zone, more
The variance of this new barrel, and the bucket of variance maximum in current variance optimization histogram is selected as bucket to be divided, the minimum bucket of variance
And its adjacent less bucket of variance is as two buckets to be combined, if data variance in two buckets to be combined and being less than
The variance of data in bucket to be divided, then perform split degree operation, conversely, not performing any operation then.
It is preferred that, it is described dynamically to update the variance optimization histogram using the new data that add, including:
The variance built in set memory optimizes histogrammic time window, if newly arrived number in the sample space of internal memory
Exceed the time window according to the corresponding sampling time, then compress the variance optimization histogram in the time window.
Further, the variance optimization histogram in the described pair of time window is compressed, including:
Retain the data boundary of each bucket in the variance optimization histogram in the time window, delete data boundary in each bucket
Outer data, wherein, the data boundary of each bucket includes:Minimal sampling time number corresponding with the maximum sampling time in each bucket
The average value of all data according to this and in this barrel.
A kind of variance optimization histogram construction device based on Spark Streaming, it is theed improvement is that, the dress
Put including:
Sampling module, for carrying out on-line sampling using Spark Streaming stream datas;
Module is built, for according to on-line sampling data, dynamic construction variance optimization histogram;
Update module, for dynamically updating the variance optimization histogram using the new data that add, and adds number according to new
Optimize histogram according to dynamic construction variance again.
It is preferred that, the sampling module, including:
RDD time interval is set, is converted to flow data according to time interval using Spark Streaming
DStream structures, and window () operating parameter of RDD in DStream structures is called, RDD in DStream structures is polymerize
For window, then on-line sampling is carried out to the data in each window;
Wherein, the time interval of window is the positive integer times of RDD time interval.
Further, the data in each window carry out on-line sampling, including:
At the beginning of the sample data of the sample space of window is set as into K, and will sample threshold value T, variable sum and variable count
Beginning turns to 0, and greatest member number is initialized for K small top pile structure;
Data in the sample space of data storage in window to window, and in window are updated into the sampling threshold
Value T;
Wherein, when the sample data of the sample space of window is K+1, selected from the sample data of the sample space of window
Select the sample data ω for meeting constraintsiAnd ωj, and by sample data ωiAnd ωjThe small data accumulation of middle data value is to number
According to being worth in big data, the sample data ω is deletediAnd ωjThe small data of middle data value.
Further, the constraints includes:Sample data ωiWith ωjSum is less than the sampling threshold value T, and sample
Data ωiAnd ωjSampling cost it is minimum in the sample space of window;
Wherein, sample data ω is determined as the following formulaiAnd ωjSampling cost J:
In above formula, ωiFor i-th of sample data, ωjFor j-th of sample data, i<j;
Further, the data in window update the sampling threshold value T, including:
Judge whether the data value in window is less than sampling threshold value T;
If so, then the data value is added on variable sum, 1 is increased certainly with variations per hour count;
If it is not, then by the data storage into the small top pile structure;
Wherein, when the data amount check in small top pile structure reaches K, or the heap of small top pile structure serves as a fill-in the value of evidence and is less than and works as
During preceding sampling threshold value T, then the value that heap serves as a fill-in evidence is added on variable sum, with variations per hour count from increasing 1, and by the heap top
Data are deleted, and the value for updating sampling threshold value T is sum/count.
It is preferred that, the structure module, including:
The maximum sample number that the sample space of set memory is allowed is KmaxAnd according to time interval setting variance optimization
The number of histogrammic bucket is B;
With the K first reached in the sample space of internal memorymaxThe data of individual online acquisition build the wide Nogata that bucket number is B
Figure, wherein, it is right to time interval institute where it to add the data of online acquisition according to the acquisition time of the data of online acquisition
The variance answered optimizes in histogrammic bucket, and determines the variance of each bucket using the data value in each bucket, builds histogram.
Further, when the data of the sample space of internal memory are KmaxAt+1, according to the sampling time of the data newly added
It is added into the variance to where it corresponding to time interval to optimize in histogrammic bucket, while updating the variance of this barrel, and selects
The bucket of variance maximum in current variance optimization histogram is selected as bucket to be divided, the minimum bucket of variance and its adjacent variance are smaller
A bucket as two buckets to be combined, if the variance of total data in two buckets to be combined is less than data in bucket to be divided
Variance, then perform split degree operation, conversely, not performing any operation then;
When the data of the sample space of internal memory are Kmax+ 1, and the data newly added sampling time not variance optimize
In the time zone of histogrammic bucket, then add it in the corresponding bucket of its sampling time nearest time zone, more
The variance of this new barrel, and the bucket of variance maximum in current variance optimization histogram is selected as bucket to be divided, the minimum bucket of variance
And its adjacent less bucket of variance is as two buckets to be combined, if data variance in two buckets to be combined and being less than
The variance of data in bucket to be divided, then perform split degree operation, conversely, not performing any operation then.
It is preferred that, the update module, including:
The variance built in set memory optimizes histogrammic time window, if newly arrived number in the sample space of internal memory
Exceed the time window according to the corresponding sampling time, then compress the variance optimization histogram in the time window.
Further, the variance optimization histogram in the described pair of time window is compressed, including:
Retain the data boundary of each bucket in the variance optimization histogram in the time window, delete data boundary in each bucket
Outer data, wherein, the data boundary of each bucket includes:Minimal sampling time number corresponding with the maximum sampling time in each bucket
The average value of all data according to this and in this barrel.
With immediate prior art ratio, the technical scheme that the present invention is provided has the advantages that:
The technical scheme that the present invention is provided, solving existing method can not enter to the magnanimity of Unknown Distribution, high speed flow data
The problem of efficient variance optimization histogram of row is built, polymerization that can be interval when limiting support high-precision real in memory headroom is looked into
Ask, not only combine Spark Streaming height and handle up, low delay, support fault-tolerant characteristic, while designed method is to grasp
The form for making operator is added in Spark Streaming calculating platforms, efficiently solves the structure side under distributed environment
The histogrammic problem of difference optimization, by RDD conversion operations new under Spark Streaming, is also achieved in linear session
Data in setting window are carried out with the optimization sampling of online variance;By in internal memory dynamic construction variance optimize it is histogrammic
Method, variance optimization histogram can be built without complicated dynamic programming algorithm limiting in space;The technology that the present invention is provided
Scheme realizes the variance under high speed magnanimity flow data environment and optimizes histogrammic Dynamic Maintenance and support interval in real time gather
Inquiry is closed, with Spark Streaming height is fault-tolerant, low delay characteristic.
Brief description of the drawings
Fig. 1 is the flow chart that a kind of variance based on Spark Streaming of the present invention optimizes histogram construction method;
Fig. 2 is the overall flow that a kind of variance based on Spark Streaming of the present invention optimizes histogram construction method
Figure;
Fig. 3 is using Spark Streaming stream datas to carry out online variance optimization sample streams in the embodiment of the present invention
Cheng Tu;
Fig. 4 is dynamic construction variance optimization histogram method flow chart in internal memory in the embodiment of the present invention;
Fig. 5 is the structural representation that a kind of variance based on Spark Streaming of the present invention optimizes histogram construction device
Figure.
Embodiment
The embodiment to the present invention elaborates below in conjunction with the accompanying drawings.
To make the purpose, technical scheme and advantage of the embodiment of the present invention clearer, below in conjunction with the embodiment of the present invention
In accompanying drawing, the technical scheme in the embodiment of the present invention is clearly and completely described, it is clear that described embodiment is
A part of embodiment of the present invention, rather than whole embodiments.Based on the embodiment in the present invention, those of ordinary skill in the art
The all other embodiment obtained under the premise of creative work is not made, belongs to the scope of protection of the invention.
With the explosive growth of current network data, the real-time analytical technology of stream data becomes the popular neck of research
Domain.Variance optimization histogram can return to the interval polymerization result of high accuracy limiting in space, have in data statistics, analysis field
Extremely important with being widely applied.But, because variance optimizes the complexity that histogram is built, currently without a kind of flow data
Computing system can provide efficient online variance optimization histogram construction method, and therefore, the present invention devises one kind and is based on
Spark Streaming variance optimization histogram construction method, is realized under conditions of memory headroom is limited, to Unknown Distribution
Out of order flow data carry out that the optimization of efficient variance is histogrammic to be built, and optimize histogram construction method and existing with traditional variance
Some flow data approximating variances optimization histogram construction methods are contrasted, test result indicates that, this method is limiting internal memory sky
It is interior that high-precision approximating variances optimization histogram can be built by single pass data, solve existing flow data and calculate
System can not efficiently build variance and optimize histogrammic problem, and main contents of the present invention include three parts:1. in Spark
Under Streaming environment, the data in DStream are carried out with online variance optimization and is sampled;2. sampled in each window of persistence
Sample data afterwards uses sample data excellent with merging structure approximating variances by the Dynamic Division of histogram bucket into internal memory
Change histogram;3. variance optimizes histogrammic Dynamic Maintenance in internal memory, and supports interval aggregate query in real time, as shown in figure 1, bag
Include:
101. carry out on-line sampling using Spark Streaming stream datas;
102. according to on-line sampling data, dynamic construction variance optimization histogram;
103. dynamically updating the variance optimization histogram using the new data that add, and data dynamic again is added according to new
Variance optimization histogram is built, i.e., returning to the step 102 according to new addition data, state builds variance optimization histogram again.
Specifically, the step 101, including:
RDD time interval is set, is converted to flow data according to time interval using Spark Streaming
DStream structures, and window () operating parameter of RDD in DStream structures is called, RDD in DStream structures is polymerize
For window, on-line sampling further is carried out to the data in each window;
Wherein, the time interval of window is the positive integer times of RDD time interval.
The data in each window carry out on-line sampling, including:
The sample data of sample space of window is set as K, initialization sampling threshold value T, variable sum and variable count are
0, and initialize the small top pile structure that greatest member number is K;
Data in the sample space of data storage in window to window, and in window are updated into the sampling threshold
Value T;
Wherein, when the sample data of the sample space of window is K+1, selected from the sample data of the sample space of window
Select the sample data ω for meeting constraintsiAnd ωj, and by sample data ωiAnd ωjThe small data accumulation of middle data value is to number
According to being worth in big data, while deleting the sample data ωiAnd ωjThe small data of middle data value.
The constraints includes:Sample data ωiWith ωjSum is less than the sampling threshold value T, and sample data ωiWith
ωjSampling cost it is minimum in the sample space of window;
Wherein, sample data ω is determined as the following formulaiAnd ωjSampling cost J:
In above formula, ωiFor i-th of sample data, ωjFor j-th of sample data, i<j;
The data in window update the sampling threshold value T, including:
Judge whether the data value in window is less than sampling threshold value T;
If so, then the data value is added on variable sum, 1 is increased certainly with variations per hour count;
If it is not, then by the data storage into the small top pile structure;
Wherein, when the data amount check in small top pile structure reaches K, or the heap of small top pile structure serves as a fill-in the value of evidence and is less than and works as
During preceding sampling threshold value T, then the value that heap serves as a fill-in evidence is added on variable sum, with variations per hour count from increasing 1, and by the heap top
Data are deleted, and the value for updating sampling threshold value T is sum/count.
The step 102, including:
The maximum sample number that the sample space of set memory is allowed is KmaxAnd according to time interval setting variance optimization
The number of histogrammic bucket is B;
Utilize the K first reached in the sample space of internal memorymaxIt is the wide straight of B that the data of individual online acquisition, which build bucket number,
Fang Tu, wherein, the data of online acquisition are added to time interval institute where it according to the acquisition time of the data of online acquisition
Corresponding variance optimizes in histogrammic bucket, and determines the variance of each bucket using the data value in each bucket, builds histogram.
Wherein, when the data of the sample space of internal memory are KmaxAt+1, according to the data newly added sampling time by its
Add to the variance corresponding to time interval where it and optimize in histogrammic bucket, while updating the variance of this barrel, and select to work as
The maximum bucket of variance is used as bucket to be divided, the minimum bucket of variance and its adjacent variance less one in the difference optimization histogram of front
Individual bucket is as two buckets to be combined, if the variance of the total data in two buckets to be combined is less than the side of data in bucket to be divided
Difference, then perform split degree operation, if data variance and not less than data in bucket to be divided the side in two buckets to be combined
Difference, then without any operation;
When the data of the sample space of internal memory are Kmax+ 1, and the data newly added sampling time not variance optimize
In the time zone of histogrammic bucket, then add it in the corresponding bucket of its sampling time nearest time zone, more
The variance of this new barrel, and the bucket of variance maximum in current variance optimization histogram is selected as bucket to be divided, the minimum bucket of variance
And its adjacent less bucket of variance is as two buckets to be combined, if data variance in two buckets to be combined and being less than
The variance of data in bucket to be divided, then perform split degree operation, if data variance in two buckets to be combined and being not less than
The variance of data in bucket to be divided, then without any operation.
In the step 103, the variance optimization histogram is dynamically updated using the new data that add, including:
The variance built in set memory optimizes histogrammic time window, if newly arrived number in the sample space of internal memory
Exceed the time window according to the corresponding sampling time, then the variance optimization histogram in the time window is compressed.
Wherein, the variance optimization histogram in the described pair of time window is compressed, including:
Retain the data boundary of each bucket in the variance optimization histogram in the time window, delete and number of boundary is removed in each bucket
According to outer data, wherein, the data boundary of each bucket includes:Minimal sampling time is corresponding with the maximum sampling time in each bucket
The average value of all data in data and this barrel.
A kind of variance optimization histogram construction device based on Spark Streaming, as shown in figure 5, described device bag
Include:
Sampling module, for carrying out on-line sampling using Spark Streaming stream datas;
Module is built, for according to on-line sampling data, dynamic construction variance optimization histogram;
Update module, for dynamically updating the variance optimization histogram using the new data that add, and adds number according to new
Optimize histogram according to dynamic construction variance again.
The sampling module, including:
RDD time interval is set, is converted to flow data according to time interval using Spark Streaming
DStream structures, and window () operating parameter of RDD in DStream structures is called, RDD in DStream structures is polymerize
For window, on-line sampling further is carried out to the data in each window;
Wherein, the time interval of window is the positive integer times of RDD time interval.
The data in each window carry out on-line sampling, including:
The sample data of sample space of window is set as K, initialization sampling threshold value T, variable sum and variable count are
0, and initialize the small top pile structure that greatest member number is K;
Data in the sample space of data storage in window to window, and in window are updated into the sampling threshold
Value T;
Wherein, when the sample data of the sample space of window is K+1, selected from the sample data of the sample space of window
Select the sample data ω for meeting constraintsiAnd ωj, and by sample data ωiAnd ωjThe small data accumulation of middle data value is to number
According to being worth in big data, while deleting the sample data ωiAnd ωjThe small data of middle data value.
The constraints includes:Sample data ωiWith ωjSum is less than the sampling threshold value T, and sample data ωiWith
ωjSampling cost it is minimum in the sample space of window;
Wherein, sample data ω is determined as the following formulaiAnd ωjSampling cost J:
In above formula, ωiFor i-th of sample data, ωjFor j-th of sample data, i<j;
The data in window update the sampling threshold value T, including:
Judge whether the data value in window is less than sampling threshold value T;
If so, then the data value is added on variable sum, 1 is increased certainly with variations per hour count;
If it is not, then by the data storage into the small top pile structure;
Wherein, when the data amount check in small top pile structure reaches K, or the heap of small top pile structure serves as a fill-in the value of evidence and is less than and works as
During preceding sampling threshold value T, then the value that heap serves as a fill-in evidence is added on variable sum, with variations per hour count from increasing 1, and by the heap top
Data are deleted, and the value for updating sampling threshold value T is sum/count.
The structure module, including:
The maximum sample number that the sample space of set memory is allowed is KmaxAnd according to time interval setting variance optimization
The number of histogrammic bucket is B;
Utilize the K first reached in the sample space of internal memorymaxIt is the wide straight of B that the data of individual online acquisition, which build bucket number,
Fang Tu, wherein, the data of online acquisition are added to time interval institute where it according to the acquisition time of the data of online acquisition
Corresponding variance optimizes in histogrammic bucket, and determines the variance of each bucket using the data value in each bucket, builds histogram.
Wherein, when the data of the sample space of internal memory are KmaxAt+1, according to the data newly added sampling time by its
Add to the variance corresponding to time interval where it and optimize in histogrammic bucket, while updating the variance of this barrel, and select to work as
The maximum bucket of variance is used as bucket to be divided, the minimum bucket of variance and its adjacent variance less one in the difference optimization histogram of front
Individual bucket is as two buckets to be combined, if the variance of the total data in two buckets to be combined is less than the side of data in bucket to be divided
Difference, then perform split degree operation, if data variance and not less than data in bucket to be divided the side in two buckets to be combined
Difference, then without any operation;
When the data of the sample space of internal memory are Kmax+ 1, and the data newly added sampling time not variance optimize
In the time zone of histogrammic bucket, then add it in the corresponding bucket of its sampling time nearest time zone, more
The variance of this new barrel, and the bucket of variance maximum in current variance optimization histogram is selected as bucket to be divided, the minimum bucket of variance
And its adjacent less bucket of variance is as two buckets to be combined, if data variance in two buckets to be combined and being less than
The variance of data in bucket to be divided, then perform split degree operation, if data variance in two buckets to be combined and being not less than
The variance of data in bucket to be divided, then without any operation.
The update module, including:
The variance built in set memory optimizes histogrammic time window, if newly arrived number in the sample space of internal memory
Exceed the time window according to the corresponding sampling time, then the variance optimization histogram in the time window is compressed.
Wherein, the variance optimization histogram in the described pair of time window is compressed, including:
Retain the data boundary of each bucket in the variance optimization histogram in the time window, delete and number of boundary is removed in each bucket
According to outer data, wherein, the data boundary of each bucket includes:Minimal sampling time is corresponding with the maximum sampling time in each bucket
The average value of all data in data and this barrel.
It should be understood by those skilled in the art that, embodiments herein can be provided as method, system or computer program
Product.Therefore, the application can be using the reality in terms of complete hardware embodiment, complete software embodiment or combination software and hardware
Apply the form of example.Moreover, the application can be used in one or more computers for wherein including computer usable program code
The computer program production that usable storage medium is implemented on (including but is not limited to magnetic disk storage, CD-ROM, optical memory etc.)
The form of product.
The application is the flow with reference to method, equipment (system) and computer program product according to the embodiment of the present application
Figure and/or block diagram are described.It should be understood that can be by every first-class in computer program instructions implementation process figure and/or block diagram
Journey and/or the flow in square frame and flow chart and/or block diagram and/or the combination of square frame.These computer programs can be provided
The processor of all-purpose computer, special-purpose computer, Embedded Processor or other programmable data processing devices is instructed to produce
A raw machine so that produced by the instruction of computer or the computing device of other programmable data processing devices for real
The device for the function of being specified in present one flow of flow chart or one square frame of multiple flows and/or block diagram or multiple square frames.
These computer program instructions, which may be alternatively stored in, can guide computer or other programmable data processing devices with spy
Determine in the computer-readable memory that mode works so that the instruction being stored in the computer-readable memory, which is produced, to be included referring to
Make the manufacture of device, the command device realize in one flow of flow chart or multiple flows and/or one square frame of block diagram or
The function of being specified in multiple square frames.
These computer program instructions can be also loaded into computer or other programmable data processing devices so that in meter
Series of operation steps is performed on calculation machine or other programmable devices to produce computer implemented processing, thus in computer or
The instruction performed on other programmable devices is provided for realizing in one flow of flow chart or multiple flows and/or block diagram one
The step of function of being specified in individual square frame or multiple square frames.
Finally it should be noted that:The above embodiments are merely illustrative of the technical scheme of the present invention and are not intended to be limiting thereof, to the greatest extent
The present invention is described in detail with reference to above-described embodiment for pipe, those of ordinary skills in the art should understand that:Still
The embodiment of the present invention can be modified or equivalent substitution, and without departing from any of spirit and scope of the invention
Modification or equivalent substitution, it all should cover within the claims of the present invention.
Claims (10)
1. a kind of variance optimization histogram construction method based on Spark Streaming, it is characterised in that methods described bag
Include:
On-line sampling is carried out using Spark Streaming stream datas;
According to on-line sampling data, dynamic construction variance optimization histogram;
The variance optimization histogram is dynamically updated using the new data that add, and data dynamic construction variance again is added according to new
Optimize histogram.
2. the method as described in claim 1, it is characterised in that described to be carried out using Spark Streaming stream datas
Line is sampled, including:
RDD time interval is set, flow data is converted into DStream knots according to time interval using Spark Streaming
Structure, and window () operating parameter of RDD in DStream structures is called, the RDD in DStream structures is polymerized to window,
On-line sampling is carried out to the data in each window again;
Wherein, the time interval of window is the positive integer times of RDD time interval.
3. method as claimed in claim 2, it is characterised in that the data in each window carry out on-line sampling, including:
The sample data of the sample space of window is set as K, and will sampling threshold value T, variable sum and variable count initialization
For 0, and greatest member number is initialized for K small top pile structure;
Data in the sample space of data storage in window to window, and in window are updated into the sampling threshold value T;
Wherein, when the sample data of the sample space of window is K+1, select full from the sample data of the sample space of window
The sample data ω of sufficient constraintsiAnd ωj, and by sample data ωiAnd ωjThe small data accumulation of middle data value is to data value
In big data, the sample data ω is deletediAnd ωjThe small data of middle data value;
It is preferred that, the constraints includes:Sample data ωiWith ωjSum is less than the sampling threshold value T, and sample data
ωiAnd ωjSampling cost it is minimum in the sample space of window;
Wherein, sample data ω is determined as the following formulaiAnd ωjSampling cost J:
<mrow>
<mi>J</mi>
<mo>=</mo>
<msub>
<mi>&omega;</mi>
<mi>i</mi>
</msub>
<mo>&CenterDot;</mo>
<msub>
<mi>&omega;</mi>
<mi>j</mi>
</msub>
<mrow>
<mo>(</mo>
<mn>1</mn>
<mo>/</mo>
<mn>3</mn>
<mo>&CenterDot;</mo>
<mo>(</mo>
<mrow>
<msub>
<mi>&omega;</mi>
<mi>i</mi>
</msub>
<mo>+</mo>
<msub>
<mi>&omega;</mi>
<mi>j</mi>
</msub>
</mrow>
<mo>)</mo>
<mo>+</mo>
<msubsup>
<mi>&Sigma;</mi>
<mrow>
<mi>m</mi>
<mo>=</mo>
<mi>i</mi>
<mo>+</mo>
<mn>1</mn>
</mrow>
<mrow>
<mi>j</mi>
<mo>-</mo>
<mn>1</mn>
</mrow>
</msubsup>
<msub>
<mi>&omega;</mi>
<mi>m</mi>
</msub>
<mo>)</mo>
</mrow>
</mrow>
In above formula, ωiFor i-th of sample data, ωjFor j-th of sample data, i<j;
It is preferred that, the data in window update the sampling threshold value T, including:
Judge whether the data value in window is less than sampling threshold value T;
If so, then the data value is added on variable sum, 1 is increased certainly with variations per hour count;
If it is not, then by the data storage into the small top pile structure;
Wherein, when the data amount check in small top pile structure reaches that K, or the value of the useful evidence of heap of small top pile structure are less than currently
During sampling threshold value T, then the value that heap serves as a fill-in evidence is added on variable sum, with variations per hour count from increasing 1, and the heap is served as a fill-in into evidence
Delete, the value for the threshold value T that samples is updated to sum/count.
4. the method as described in claim 1, it is characterised in that the data according to on-line sampling, dynamic construction variance is excellent
Change histogram, including:
The maximum sample number that the sample space of set memory is allowed is KmaxAnd according to time interval setting variance optimization Nogata
The number of the bucket of figure is B;
With the K first reached in the sample space of internal memorymaxThe data of individual online acquisition build the wide histogram that bucket number is B, its
In, the data of online acquisition are added to the side corresponding to time interval where it according to the acquisition time of the data of online acquisition
Difference optimizes in histogrammic bucket, and determines the variance of each bucket using the data value in each bucket, builds histogram;
It is preferred that, when the data of the sample space of internal memory are KmaxAt+1, added according to the sampling time of the data newly added
Enter to the variance corresponding to time interval where it and optimize in histogrammic bucket, while updating the variance of this barrel, and select current
The maximum bucket of variance is used as bucket to be divided, the minimum bucket of variance and its less one adjacent of variance in variance optimization histogram
Bucket is as two buckets to be combined, if the variance of the total data in two buckets to be combined is less than the side of data in bucket to be divided
Difference, then perform split degree operation, conversely, not performing any operation then;
When the data of the sample space of internal memory are Kmax+ 1, and the data newly added sampling time not variance optimize Nogata
In the time zone of the bucket of figure, then add it in the corresponding bucket of its sampling time nearest time zone, updating should
The variance of bucket, and select in current variance optimization histogram the maximum bucket of variance as bucket divide, the bucket of variance minimum and its
The less bucket of adjacent variance is as two buckets to be combined, if data variance in two buckets to be combined and less than treating point
The variance of data in bucket is split, then split degree operation is performed, conversely, not performing any operation then.
5. the method as described in claim 1, it is characterised in that described dynamically to update the variance optimization using the new data that add
Histogram, including:
The variance built in set memory optimizes histogrammic time window, if newly arrived data pair in the sample space of internal memory
The sampling time answered exceeds the time window, then compresses the variance optimization histogram in the time window;
It is preferred that, the variance optimization histogram compressed in the time window, including:
Retain the data boundary of each bucket in the variance optimization histogram in the time window, delete in each bucket outside data boundary
Data, wherein, the data boundary of each bucket includes:In each bucket minimal sampling time data corresponding with the maximum sampling time with
And in this barrel all data average value.
6. a kind of variance optimization histogram construction device based on Spark Streaming, it is characterised in that described device bag
Include:
Sampling module, for carrying out on-line sampling using Spark Streaming stream datas;
Module is built, for according to on-line sampling data, dynamic construction variance optimization histogram;
Update module, for dynamically updating the variance optimization histogram using the new data that add, and adds data weight according to new
New dynamic construction variance optimization histogram.
7. device as claimed in claim 6, it is characterised in that the sampling module, including:
RDD time interval is set, flow data is converted into DStream knots according to time interval using Spark Streaming
Structure, and window () operating parameter of RDD in DStream structures is called, RDD in DStream structures is polymerized to window, then
On-line sampling is carried out to the data in each window;
Wherein, the time interval of window is the positive integer times of RDD time interval.
8. device as claimed in claim 7, it is characterised in that the data in each window carry out on-line sampling, including:
The sample data of the sample space of window is set as K, and will sampling threshold value T, variable sum and variable count initialization
For 0, and greatest member number is initialized for K small top pile structure;
Data in the sample space of data storage in window to window, and in window are updated into the sampling threshold value T;
Wherein, when the sample data of the sample space of window is K+1, select full from the sample data of the sample space of window
The sample data ω of sufficient constraintsiAnd ωj, and by sample data ωiAnd ωjThe small data accumulation of middle data value is to data value
In big data, the sample data ω is deletediAnd ωjThe small data of middle data value;
It is preferred that, the constraints includes:Sample data ωiWith ωjSum is less than the sampling threshold value T, and sample data
ωiAnd ωjSampling cost it is minimum in the sample space of window;
Wherein, sample data ω is determined as the following formulaiAnd ωjSampling cost J:
<mrow>
<mi>J</mi>
<mo>=</mo>
<msub>
<mi>&omega;</mi>
<mi>i</mi>
</msub>
<mo>&CenterDot;</mo>
<msub>
<mi>&omega;</mi>
<mi>j</mi>
</msub>
<mrow>
<mo>(</mo>
<mn>1</mn>
<mo>/</mo>
<mn>3</mn>
<mo>&CenterDot;</mo>
<mo>(</mo>
<mrow>
<msub>
<mi>&omega;</mi>
<mi>i</mi>
</msub>
<mo>+</mo>
<msub>
<mi>&omega;</mi>
<mi>j</mi>
</msub>
</mrow>
<mo>)</mo>
<mo>+</mo>
<msubsup>
<mi>&Sigma;</mi>
<mrow>
<mi>m</mi>
<mo>=</mo>
<mi>i</mi>
<mo>+</mo>
<mn>1</mn>
</mrow>
<mrow>
<mi>j</mi>
<mo>-</mo>
<mn>1</mn>
</mrow>
</msubsup>
<msub>
<mi>&omega;</mi>
<mi>m</mi>
</msub>
<mo>)</mo>
</mrow>
</mrow>
In above formula, ωiFor i-th of sample data, ωjFor j-th of sample data, i<j;
It is preferred that, the data in window update the sampling threshold value T, including:
Judge whether the data value in window is less than sampling threshold value T;
If so, then the data value is added on variable sum, 1 is increased certainly with variations per hour count;
If it is not, then by the data storage into the small top pile structure;
Wherein, when the data amount check in small top pile structure reaches that K, or the value of the useful evidence of heap of small top pile structure are less than currently
During sampling threshold value T, then the value that heap serves as a fill-in evidence is added on variable sum, with variations per hour count from increasing 1, and the heap is served as a fill-in into evidence
Delete, the value for updating sampling threshold value T is sum/count.
9. device as claimed in claim 6, it is characterised in that the structure module, including:
The maximum sample number that the sample space of set memory is allowed is KmaxAnd according to time interval setting variance optimization Nogata
The number of the bucket of figure is B;
With the K first reached in the sample space of internal memorymaxThe data of individual online acquisition build the wide histogram that bucket number is B, its
In, the data of online acquisition are added to the side corresponding to time interval where it according to the acquisition time of the data of online acquisition
Difference optimizes in histogrammic bucket, and determines the variance of each bucket using the data value in each bucket, builds histogram;
It is preferred that, when the data of the sample space of internal memory are KmaxAt+1, added according to the sampling time of the data newly added
Enter to the variance corresponding to time interval where it and optimize in histogrammic bucket, while updating the variance of this barrel, and select current
The maximum bucket of variance is used as bucket to be divided, the minimum bucket of variance and its less one adjacent of variance in variance optimization histogram
Bucket is as two buckets to be combined, if the variance of the total data in two buckets to be combined is less than the side of data in bucket to be divided
Difference, then perform split degree operation, conversely, not performing any operation then;
When the data of the sample space of internal memory are Kmax+ 1, and the data newly added sampling time not variance optimize Nogata
In the time zone of the bucket of figure, then add it in the corresponding bucket of its sampling time nearest time zone, updating should
The variance of bucket, and select in current variance optimization histogram the maximum bucket of variance as bucket divide, the bucket of variance minimum and its
The less bucket of adjacent variance is as two buckets to be combined, if data variance in two buckets to be combined and less than treating point
The variance of data in bucket is split, then split degree operation is performed, conversely, not performing any operation then.
10. device as claimed in claim 6, it is characterised in that the update module, including:
The variance built in set memory optimizes histogrammic time window, if newly arrived data pair in the sample space of internal memory
The sampling time answered exceeds the time window, then compresses the variance optimization histogram in the time window;
It is preferred that, the variance optimization histogram compressed in the time window, including:
Retain the data boundary of each bucket in the variance optimization histogram in the time window, delete in each bucket outside data boundary
Data, wherein, the data boundary of each bucket includes:In each bucket minimal sampling time data corresponding with the maximum sampling time with
And in this barrel all data average value.
Priority Applications (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201710212747.7A CN107193862A (en) | 2017-04-01 | 2017-04-01 | A kind of variance optimization histogram construction method and device based on Spark Streaming |
Applications Claiming Priority (1)
Application Number | Priority Date | Filing Date | Title |
---|---|---|---|
CN201710212747.7A CN107193862A (en) | 2017-04-01 | 2017-04-01 | A kind of variance optimization histogram construction method and device based on Spark Streaming |
Publications (1)
Publication Number | Publication Date |
---|---|
CN107193862A true CN107193862A (en) | 2017-09-22 |
Family
ID=59871263
Family Applications (1)
Application Number | Title | Priority Date | Filing Date |
---|---|---|---|
CN201710212747.7A Pending CN107193862A (en) | 2017-04-01 | 2017-04-01 | A kind of variance optimization histogram construction method and device based on Spark Streaming |
Country Status (1)
Country | Link |
---|---|
CN (1) | CN107193862A (en) |
Cited By (1)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110362600A (en) * | 2019-07-22 | 2019-10-22 | 广西大学 | A kind of random ordering data flow distribution aggregate query method, system and medium |
Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102103638A (en) * | 2010-02-01 | 2011-06-22 | 北京大学 | Query log-based database statistic data histogram generation method |
US20120102377A1 (en) * | 2010-10-26 | 2012-04-26 | Krishnamurthy Viswanathan | Method for constructing a histogram |
CN104657450A (en) * | 2015-02-05 | 2015-05-27 | 中国科学院信息工程研究所 | Big data environment-oriented summary information dynamic constructing and querying method and device |
CN105718872A (en) * | 2016-01-15 | 2016-06-29 | 武汉光庭科技有限公司 | Auxiliary method and system for rapid positioning of two-side lanes and detection of deflection angle of vehicle |
-
2017
- 2017-04-01 CN CN201710212747.7A patent/CN107193862A/en active Pending
Patent Citations (4)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN102103638A (en) * | 2010-02-01 | 2011-06-22 | 北京大学 | Query log-based database statistic data histogram generation method |
US20120102377A1 (en) * | 2010-10-26 | 2012-04-26 | Krishnamurthy Viswanathan | Method for constructing a histogram |
CN104657450A (en) * | 2015-02-05 | 2015-05-27 | 中国科学院信息工程研究所 | Big data environment-oriented summary information dynamic constructing and querying method and device |
CN105718872A (en) * | 2016-01-15 | 2016-06-29 | 武汉光庭科技有限公司 | Auxiliary method and system for rapid positioning of two-side lanes and detection of deflection angle of vehicle |
Non-Patent Citations (4)
Title |
---|
夏小玲 等: "面向数据流的差分隐私直方图发布", 《计算机与现代化》 * |
杨颖 等: "动态地构建和维护基于小波的直方图", 《计算机应用研究》 * |
林子雨老师: "Spark入门:Spark Streaming简介", 《HTTP://DBLAB.XMU.EDU.CN/BLOG/1076-2/》 * |
黄超 等: "时间序列数据流直方图构造方法研究", 《统计与决策》 * |
Cited By (2)
Publication number | Priority date | Publication date | Assignee | Title |
---|---|---|---|---|
CN110362600A (en) * | 2019-07-22 | 2019-10-22 | 广西大学 | A kind of random ordering data flow distribution aggregate query method, system and medium |
CN110362600B (en) * | 2019-07-22 | 2022-03-11 | 广西大学 | Out-of-order data stream distributed aggregation query method, system and medium |
Similar Documents
Publication | Publication Date | Title |
---|---|---|
CN105227488B (en) | A kind of network flow group scheduling method for distributed computer platforms | |
CN101252541B (en) | Method for establishing network flow classified model and corresponding system thereof | |
CN104376365A (en) | Method for constructing information system running rule libraries on basis of association rule mining | |
CN106777093A (en) | Skyline inquiry systems based on space time series data stream application | |
CN110222029A (en) | A kind of big data multidimensional analysis computational efficiency method for improving and system | |
CN103745273A (en) | Semiconductor fabrication process multi-performance prediction method | |
CN101000605A (en) | Intelligent two-stage compression method for process industrial historical data | |
CN106708989A (en) | Spatial time sequence data stream application-based Skyline query method | |
CN108416465A (en) | A kind of Workflow optimization method under mobile cloud environment | |
CN107644252A (en) | A kind of recurrent neural networks model compression method of more mechanism mixing | |
CN106161135A (en) | Business transaction failure analysis methods and device | |
CN114513470A (en) | Network flow control method, device, equipment and computer readable storage medium | |
CN116827350A (en) | Flexible work platform intelligent supervision method and system based on cloud edge cooperation | |
CN116050540A (en) | Self-adaptive federal edge learning method based on joint bi-dimensional user scheduling | |
CN114781650A (en) | Data processing method, device, equipment and storage medium | |
CN110263917A (en) | A kind of neural network compression method and device | |
CN107193862A (en) | A kind of variance optimization histogram construction method and device based on Spark Streaming | |
CN114169506A (en) | Deep learning edge computing system framework based on industrial Internet of things platform | |
CN110309955A (en) | A kind of non-load predicting method and device shut down when upgrading of cloud environment application system | |
CN114282658B (en) | Method, device and medium for analyzing and predicting flow sequence | |
CN106909459A (en) | A kind of method and device for adjusting connection pool | |
CN110191005A (en) | A kind of alarm log processing method and system | |
CN109800271A (en) | A kind of information collecting method based on big data | |
CN107704565A (en) | Equivalent body generation method, device and system | |
CN110989040A (en) | Artificial intelligent lightning approach early warning method and system based on slice processing |
Legal Events
Date | Code | Title | Description |
---|---|---|---|
PB01 | Publication | ||
PB01 | Publication | ||
SE01 | Entry into force of request for substantive examination | ||
SE01 | Entry into force of request for substantive examination | ||
AD01 | Patent right deemed abandoned |
Effective date of abandoning: 20240119 |
|
AD01 | Patent right deemed abandoned |