CN109920025A - A kind of mass data is sampled drawing practice and system in the high-efficiency layered of time-domain - Google Patents
A kind of mass data is sampled drawing practice and system in the high-efficiency layered of time-domain Download PDFInfo
- Publication number
- CN109920025A CN109920025A CN201910101886.1A CN201910101886A CN109920025A CN 109920025 A CN109920025 A CN 109920025A CN 201910101886 A CN201910101886 A CN 201910101886A CN 109920025 A CN109920025 A CN 109920025A
- Authority
- CN
- China
- Prior art keywords
- data
- sampling
- time
- domain
- mass data
- 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.)
- Granted
Links
- 238000005070 sampling Methods 0.000 claims abstract description 99
- 238000000034 method Methods 0.000 claims abstract description 22
- 238000004422 calculation algorithm Methods 0.000 claims abstract description 15
- 238000012163 sequencing technique Methods 0.000 claims abstract description 5
- 229940093334 flomax Drugs 0.000 claims description 9
- ZZIZZTHXZRDOFM-XFULWGLBSA-N tamsulosin hydrochloride Chemical compound [H+].[Cl-].CCOC1=CC=CC=C1OCCN[C@H](C)CC1=CC=C(OC)C(S(N)(=O)=O)=C1 ZZIZZTHXZRDOFM-XFULWGLBSA-N 0.000 claims description 9
- 238000004364 calculation method Methods 0.000 claims description 7
- 238000004590 computer program Methods 0.000 claims description 5
- 239000000284 extract Substances 0.000 claims description 4
- 238000005086 pumping Methods 0.000 claims 1
- 238000007405 data analysis Methods 0.000 abstract description 3
- 230000035939 shock Effects 0.000 description 7
- 238000012360 testing method Methods 0.000 description 7
- 230000008569 process Effects 0.000 description 5
- 238000005516 engineering process Methods 0.000 description 4
- 230000000052 comparative effect Effects 0.000 description 3
- 238000012545 processing Methods 0.000 description 3
- 238000004458 analytical method Methods 0.000 description 2
- 230000003139 buffering effect Effects 0.000 description 2
- 230000006870 function Effects 0.000 description 2
- 239000007787 solid Substances 0.000 description 2
- 230000000007 visual effect Effects 0.000 description 2
- 241001269238 Data Species 0.000 description 1
- 240000008042 Zea mays Species 0.000 description 1
- 235000005824 Zea mays ssp. parviglumis Nutrition 0.000 description 1
- 235000002017 Zea mays subsp mays Nutrition 0.000 description 1
- 230000005540 biological transmission Effects 0.000 description 1
- 238000009933 burial Methods 0.000 description 1
- 235000005822 corn Nutrition 0.000 description 1
- 238000007418 data mining Methods 0.000 description 1
- 238000013500 data storage Methods 0.000 description 1
- 230000007812 deficiency Effects 0.000 description 1
- 230000001419 dependent effect Effects 0.000 description 1
- 238000013461 design Methods 0.000 description 1
- 238000011161 development Methods 0.000 description 1
- 230000000763 evoking effect Effects 0.000 description 1
- 230000001939 inductive effect Effects 0.000 description 1
- 238000012986 modification Methods 0.000 description 1
- 230000004048 modification Effects 0.000 description 1
- 230000003287 optical effect Effects 0.000 description 1
- 239000013307 optical fiber Substances 0.000 description 1
- 238000005457 optimization Methods 0.000 description 1
- 230000001737 promoting effect Effects 0.000 description 1
- 230000010349 pulsation Effects 0.000 description 1
- 230000009467 reduction Effects 0.000 description 1
- 238000012827 research and development Methods 0.000 description 1
- 239000004065 semiconductor Substances 0.000 description 1
Abstract
The invention belongs to data analysis technique field, discloses a kind of mass data and sample drawing practice and system in the high-efficiency layered of time-domain, mass data is sampled using high-efficiency layered sampling algorithm according to drawing window width, obtains 4X data from the sample survey；It is drawn with 4X data from the sample survey, obtains full datagraphic.The present invention provides a kind of mass data and samples drawing practice in the high-efficiency layered of time-domain, evidence totally is divided into 2X subset according to drawing window width, in each subset all data seek respectively maximum and minimum value and by the sequencing of appearance store, 4X obtained data from the sample survey figure will with it is totally consistent according to pattern height；It is compared totally according to plotting speed, " single thread sampling+drawing " bulk velocity promotes about 70 times, and " thread pool sampling+drawing " bulk velocity promotes about 200 times；The invention will increase substantially mass data time-domain plotting speed, and sampling algorithm is equally applicable to other all sequences and draws.
Description
Technical field
The invention belongs to data analysis technique fields more particularly to a kind of mass data to sample in the high-efficiency layered of time-domain
Drawing practice and system.
Background technique
Currently, the prior art commonly used in the trade is such that
With the development of informationized society, high sampling instrument used in mankind's progress scientific observation is more and more, data
It measures also increasing.The observation instrument of time-domain is many kinds of, such as: the company of seismic system sample rate 100HZ used at present
Continuous seismic waveform register instrument, continuously magnetic inductive magnetometer of sample rate 30HZ etc..When using these observation data, all
It needs to provide for user visual in image time-domain figure (horizontal axis is time scale), it must be all data by successive when drawing
Sequence line (or draw a little etc.), plotting speed become the pain place of numerous software developers and user slowly.
Plotting speed had both been decided by that hardware was also dependent upon software, can only sufficiently excavate in the case where hardware device is constant soft
The potentiality of part make software on links all with the work of maximum speed.Such as: 100HZ consecutive shock waveform instrument has three
A channel, the daily total amount of data in each channel are 8640000, and three channels are 1 day total big according to (98.88MB) non-drawing time
About 46 seconds, common double buffering graphics display technology is still utilized in current software development in this, and (property of drawing in memory again was pasted
Picture) and GraphicsPath (path drawing) in NET and multichannel thread pool (ThreadPool) draw parallel etc.
Prestissimo after technical optimization.And observation instrument as the whole nation has thousands of sets, these instruments are all continuous at 24 hours daily
Record, each office, province, which is frequently necessary to for tens sets of instruments being inside the province painted on a graphical interfaces, to be compared and analyzed and handles, because
For the figure of their slow usually each only one hour of processing of plotting speed, daily data processing at least divides 24 completions.
Why partially slow these map data speed are, and basic reason is that draw data amount is too big, and the most of the time disappears
Consumption is caused in full data bus connection (or draw a little etc.), all thoroughly can not fundamentally solve speed by existing all drawing optimisation techniques
Degree problem, only way can only be drawn using sampling.Because of horizontal axis resolution ratio usually only 2000 pixels of computer or so,
When points of drawing are considerably beyond horizontal axis definition pixel, there is a large amount of data and curves quilt behind each pixel on display
Burial is not shown at all.Problem is how sampling pattern and full datagraphic to be sampled and guaranteed to mass data
Height is consistent.
In data analysis, sampling refers to that selected section data are analyzed from total evidence, to excavate more extensive number
According to the useful information of concentration.In statistics, sampling is a kind of inferential statistics method, refers to and extracts a part from target population
Individual is used as sample, by observing a certain or certain attribute of sample, to reach to overall understanding.
There are three types of the common big data methods of samplings, (1) random sampling: randomly selecting n unit from overall N number of unit
As sample, so that each group observations in data set have the identical probability being sampled.(2) equidistant sampling: will be overall
Whole units be arranged in order, first sample unit of random sampling (or be random start), then sequentially extract remaining
Sample unit.(3) sample: being totally divided into several subsets by stratified sampling, inspects certain amount sample by random samples from each subset respectively, often
The observation that height is concentrated can set identical probability, can also set different probability, sampling results usually have better generation
Table.
Because guarantee sampling pattern with it is totally consistent according to pattern height, it is clear that random sampling and equidistant sampling all can not
With.Because the total isolated point being had at any time in similar to spike, it is at an arbitrary position and right that these isolated points are possible to occur
The configuration of figure affects greatly, both methods of samplings all not can guarantee these isolated points and accurately can be extracted into and be shown
Show on figure.
In conclusion problem of the existing technology is:
The slow basic reason of mass data plotting speed is that draw data amount is too big, and the most of the time consumes in total evidence
Caused by line (or draw a little etc.).It commonly draws in software development at present optimisation technique, as double buffering graphics display (is drawn again in memory
Disposable paste picture), the GraphicsPath (path drawing) in NET, multichannel thread pool (ThreadPool) draw parallel
Figure etc., these drawing optimisation techniques all thoroughly can not fundamentally solve speed issue.Only way can only be drawn using sampling
Figure, actually using which kind of sampling algorithm just can guarantee sampling pattern with it is totally consistent according to pattern height, become solution plotting speed
The very corn of a subject.
Solve the difficulty and meaning of above-mentioned technical problem:
(1) no matter use which kind of sampling algorithm, no matter mass data normally whether the form of mistake (even), it is necessary to
Guarantee sampling pattern and totally consistent according to pattern height, or from being visually difficult to find out the difference of the two.Its realistic meaning is, with most
Few data from the sample survey draws the full datagraphic that user wants to see, to thoroughly solve the problems, such as that mass data plotting speed is slow.
(2) no matter which kind of sampling algorithm used, it is necessary to guarantee that sampling rate is very fast, cannot have too to bulk velocity
It is big to influence.Because sampling process all carries out in memory, sampling rate is necessarily much better than plotting speed, but data volume is bigger
Sampling rate is necessarily slower when (millions of or more).Difficult point is how the influence data volume size to sampling rate reduces
To minimum.
(3) the draw draw data that uses of sampling is not total evidence, user be frequently necessary to by curve step by step partial enlargement with
See the full data information being more buried, difficult point is reflecting between the pixel of mouse picking to be accomplished and truthful data point
It penetrates, cannot have any influence to the zoom function of full datagraphic because of sampling.
Summary of the invention
In view of the problems of the existing technology, the present invention provides a kind of completely new stratified sampling drawing practices and system.
Total evidence is divided into 2X subset according to drawing window width (X pixel) by the present invention, and all data are asked respectively in each 2X subset
It takes maximum and minimum value and is stored by the sequencing of appearance, obtain 4X data from the sample survey.At this point, this 4X data from the sample survey figure
Will with it is totally consistent according to pattern height.
By taking 8640000 seismic waveform datas are drawn as an example, if drawing window width is 1000 pixels, totally according to general
It is divided into 2*1000=2000 subset, each subset data sum is 8640000/2000=4320, obtained data from the sample survey
Sum is 4*1000=4000.The same pixel of horizontal axis is all projected in each subset after 4320 data interconnecting lines
On, therefore the line of min-max (or maximum-minimum) perpendicular to horizontal axis and can cover this 4320 data in fact between them
Corresponding pixel.Although all having carried out line to this 4320 data when full map data and having spent a large amount of time,
If the line of min-max between them (or maximum-minimum) in fact, therefore a large amount of non-drawing time can be saved.
The invention is realized in this way a kind of mass data includes: in the high-efficiency layered sampling drawing practice of time-domain
Step 1 is sampled mass data using high-efficiency layered sampling algorithm according to drawing window width, obtains 4X
A data from the sample survey；
Step 2 is drawn with 4X data from the sample survey, obtains full datagraphic.
Further, the mass data is specifically included in the high-efficiency layered methods of sampling of time-domain:
The first step, draw data sum=N, drawing window width=X pixel, data from the sample survey sum=4X；
Second step, N > 4X?
Third step, if it is not, unsample, draw data=total evidence；
4th step extracts 4X data from the sample survey if so, being sampled to total evidence；
5th step, draw data=4X data from the sample survey.
Further, all data are divided into 2X subset, each 2X subset by the mass data high-efficiency layered methods of sampling
Interior all data are sought maximum and minimum value respectively and are stored by the sequencing of appearance, and 4X data from the sample survey is obtained；Specific packet
It includes:
Step 1: draw data sum=N, drawing window width=X pixel, data from the sample survey sum=4X；
Step 2:N > 4X?
Step 3: if it is not, unsample, draw data=total evidence；
Step 4: if so, class interval douJG=(double) data count N/ (2*X)；Every group of data amount check P=(takes
It is whole) douJG；
Step 5: cycle calculations grouping serial number intLen [i]=(rounding) (douJG*i), wherein i=0 ... 2X+1,
Time interval (data count for representing each subset) value of each element of intLen [i]=0 ... N, intLen are as follows: P P ... P
P+1P ... P P+1P P increases 1 every several, remaining all P；
Step 6: cycle calculations sampled value floSam [i], i=0 ... 2X-1, floSam sum are 4X；
Step 7: extracting serial number all data between intLen [i] to intLen [i+1] -1, calculate maximum value
FloMax, minimum value floMin, maximum value serial number intPosMax, minimum value serial number intPosMin；
Step 8: minimum value is formerly or maximum value is first, sequentially stores；
FloSam [2i]=intPosMin < intPosMax? floMin:floMax；
FloSam [2i+1]=intPosMin < intPosMax? floMax:floMin；
Does step 9: circulation terminate? if it is not, repeating step 6；
Step 10: if so, draw data=4X data from the sample survey.
Further, above-mentioned sampling process all carries out in memory, sampling process substantially correspond to it is totally primary according to repeating query,
Although sampling rate is far faster than according to plotting speed, sampling rate is inevitable when total amount of data is bigger (millions of or more) entirely totally
It is slower.Sampling rate can be advanced optimized with thread pool, i.e., multiple thread parallels execute sampling process, at this time sampling rate
It can obtain rapid promotion.
Further, the thinking of zoom control step by step of above-mentioned drawing includes: and calculates really to count in drawing window after sampling every time
According to the multiplying power relationship (xFdbs=X/N) between total (N) and drawing window pixel number (X).When each pattern visual evoked potentials, according to
The pixel coverage that family is selected in graphical interfaces, the truthful data range selected by xFdbs inverse store total in drawing window
According to start-stop serial number, the data chosen are sampled and are drawn again；When each pattern reduction, according to the start-stop serial number of storage
Again these data are sampled and are drawn.
Realize that the mass data is drawn in the high-efficiency layered sampling of time-domain another object of the present invention is to provide a kind of
The computer program of drawing method.
Realize that the mass data is drawn in the high-efficiency layered sampling of time-domain another object of the present invention is to provide a kind of
High-efficiency layered sampling drawing control system of the mass data of drawing method in time-domain.
In conclusion the present invention compare the prior art have it is following the utility model has the advantages that
In the identical situation of total amount of data, sampling plotting speed is much better than full map data.Plotting speed test result
Being shown in Table 2, (speed unit is the second, is abbreviated as s).
2 plotting speed test result of table
From the plotting speed test result of table 2: by taking 100HZ consecutive shock waveform instrument as an example, three, the instrument logical
Road total evidence (98.88MB) non-drawing time about 46 seconds 1 day, " single thread sampling+drawing " total time is 0.73 second, and " thread pool is taken out
Sample+drawing " total time is 0.34 second.The total evidence in 36, instrument channel 1 day (1186MB) non-drawing time about 546 seconds, " single thread
Sampling+drawing " total time is 7.77 seconds, and " thread pool sampling+drawing " total time is 2.31 seconds.Totally according to plotting speed phase
Than " single thread sampling+drawing " bulk velocity promotes about 70 times, and " thread pool sampling+drawing " bulk velocity promotes about 200 times.
The present invention draws the full datagraphic that user wants to see, sampling pattern and full datagraphic with least data from the sample survey
Height is consistent (0.99 or more similarity), to thoroughly solve the problems, such as that mass data plotting speed is slow.
Result of the present invention will significantly promote mass data service efficiency, all have to software development and the design of drawing thinking
Higher reference value, other all sequences that sampling algorithm is equally applicable to outside time-domain are drawn, to data mining and data
The fields such as analysis equally have extensive reference significance.
Detailed description of the invention
Fig. 1 is high-efficiency layered sampling drawing practice flow chart of the mass data provided in an embodiment of the present invention in time-domain.
Fig. 2 is high-efficiency layered sampling algorithm flow chart provided in an embodiment of the present invention.
Fig. 3 is high-efficiency layered sampling pattern provided in an embodiment of the present invention and totally according to graphic correlation.
Fig. 4 is that distortion phenomenon figure occurs in sampling pattern when total evidence provided in an embodiment of the present invention is divided into 1X subset.
Specific embodiment
In order to make the objectives, technical solutions, and advantages of the present invention clearer, with reference to embodiments, to the present invention
It is further elaborated.It should be appreciated that the specific embodiments described herein are merely illustrative of the present invention, it is not used to
Limit the present invention.
Mass data is that draw data amount is too big caused in the slow basic reason of time-domain plotting speed, by existing institute
There is drawing optimisation technique all thoroughly can not fundamentally solve the problems, such as that plotting speed, only way can only be drawn using sampling.
Actually using which kind of sampling algorithm just can guarantee sampling pattern with it is totally consistent according to pattern height, become and solve the problems, such as plotting speed
Core.
In order to solve the above technical problems, below with reference to concrete scheme, the present invention is described in detail.
Mass data provided in an embodiment of the present invention includes according to drawing in the high-efficiency layered sampling drawing practice of time-domain
Window width (X pixel) is sampled mass data using high-efficiency layered sampling algorithm, obtains 4X data from the sample survey；Use 4X
A data from the sample survey is drawn, obtained figure will with it is totally consistent (local minute differences) according to pattern height.
Wherein, the mass data high-efficiency layered methods of sampling includes: that all data are divided into 2X subset, each 2X subset
Interior all data are sought maximum and minimum value respectively and are stored by the sequencing that it occurs, and 4X data from the sample survey is obtained.
As shown in Figure 1, high-efficiency layered sampling drawing practice tool of the mass data provided in an embodiment of the present invention in time-domain
Body includes:
S101: draw data sum=N, drawing window width=X pixel, data from the sample survey sum=4X.
S102:N > 4X?
S103: if it is not, unsample, draw data=total evidence.
S104: if so, being sampled to total evidence, 4X data from the sample survey is extracted.
S105: draw data=4X data from the sample survey.
As shown in Fig. 2, high-efficiency layered sampling algorithm provided in an embodiment of the present invention the following steps are included:
Step 201: draw data sum=N, drawing window width=X pixel, data from the sample survey sum=4X.
Step 202:N > 4X?
Step 203: if it is not, unsample, draw data=total evidence.
Step 204: if so, class interval douJG=(double) data count N/ (2*X), douJG is necessary for double
Type guarantees that it has enough precision, otherwise may result in the last one value of intLen in step 205 and is not equal to data count
N；Every group of data amount check P=(rounding) douJG, i.e., the data amount check of average every group of distribution.
Step 205: cycle calculations grouping serial number intLen [i]=(rounding) (rounding up) (douJG*i), wherein i=
0 ... 2X+1, intLen [i]=0 ... N (the last one value is necessary for data count N), the time interval of each element of intLen
(data count for representing each subset) value are as follows: P P ... P P+1P ... P P+1P P increases 1 every several, remaining is all
P.Meaning: the class interval douJG in step 204 is usually decimal (data count N and 2X can not always divide exactly),
Remaining data number after mean allocation is evenly distributed in X-axis and (increases 1 every several) at this time.
Step 206: cycle calculations sampled value floSam [i], i=0 ... 2X-1, floSam sum are 4X.
Step 207: extracting serial number intLen [i] to all data between intLen [i+1] -1, calculate maximum value
FloMax, minimum value floMin, maximum value serial number intPosMax, minimum value serial number intPosMin.
Step 208: minimum value is formerly or maximum value is first, sequentially stores.
FloSam [2i]=intPosMin < intPosMax? floMin:floMax,
FloSam [2i+1]=intPosMin < intPosMax? floMax:floMin,
Does step 209: circulation terminate? if it is not, repeating step 206.
Step 210: if so, draw data=4X data from the sample survey.
As shown in figure 3, high-efficiency layered sampling pattern provided in an embodiment of the present invention and totally according in graphic correlation,
Fig. 3 a and Fig. 3 b are 100HZ consecutive shock waveform 1 minute (6000 data) comparative pattern.
Fig. 3 c and Fig. 3 d are 100HZ consecutive shock waveform 1 hour (360000 data) comparative pattern.
Fig. 3 e and Fig. 3 f are 100HZ consecutive shock waveform 1 day (8640000 data) comparative pattern.
Sampling pattern (2X subset) and full datagraphic have local minute differences in Fig. 3, and naked eyes do not see usual difficulty carefully
To differentiate.Figure is a reference tool after all, these minute differences are completely negligible for a user, partial enlargement
These minute differences can be disappeared automatically (with full map data when N≤4X) afterwards.
As shown in figure 4, there is distortion phenomenon in sampling pattern when total evidence provided in an embodiment of the present invention is divided into 1X subset
In, 100HZ consecutive shock waveform 10 minutes (60000 data) full datagraphics (Fig. 4 a) and total evidence are divided into 2X subset
The sampling pattern of (Fig. 4 b) and 1X subset (Fig. 4 c).Fig. 4 c causes population of samples deficiency to go out because total according to 1X subset is divided into
Existing distortion phenomenon.
In the present invention, high-efficiency layered sampling pattern and according to shape similarity calculating include: totally sampling pattern and total
It is painted in the space of same widths (W pixel) and height (H pixel) (black lines, white background) according to figure, all pictures of the two
Vegetarian refreshments is compared, and statistical color is worth identical pixel summation N, similarity=N/ (W × H).Table 1 is for sampling pattern and totally
According to the similarity calculation result of figure.
1 sampling pattern of table and totally according to shape similarity calculated result
Illustrate: drawing window width is 840 pixels (X=840), is highly 195 pixels.
Seen from table 1, total according to when being divided into 1X subset, the similarity of sampling pattern and full datagraphic is usually 0.98
Or it is declined slightly compared with 2X and 4X hereinafter, being divided into similarity when 3X or 5X.It is divided into 2X, 4X, 6X, 8X or higher (with even-multiple
It is incremented by) similarity is gradually incremented by, and when 50X, reaches 0.9999 or so, and sampling pattern and full datagraphic are substantially completely consistent at this time.
Below with reference to the test of high-efficiency layered sampling plotting speed, the invention will be further described.
Test data is 100HZ consecutive shock waveform, and the daily data count in each channel is 8640000, divides 3 to lead to
Road 1 day, 36 channels 1 day, 72 channels, 1 day three kinds of situation, plotting speed test result is shown in Table 2, and (speed unit is the second, is write a Chinese character in simplified form
For s).
2 plotting speed test result of table
Illustrate: drawing section uses Thread Pool Technology, thread of every curve (or each channel)；Always
When=sample time+non-drawing time, when promoting multiplying power=totally according to non-drawing time/total.
As can be seen from Table 2, when single thread is sampled the sample time account for the overwhelming majority of total time, and with full total amount of data
The bigger sample time is slower.After using thread pool sampling instead, sampling rate is improved significantly.Totally according to plotting speed phase
Than " single thread sampling+drawing " bulk velocity promotes about 70 times, and " thread pool sampling+drawing " bulk velocity promotes about 200 times.
Below with reference to experimental result, the invention will be further described.
It is total according to why 2X subset is divided into the high-efficiency layered sampling algorithm flow chart of Fig. 2, it is because of 2X subset
Subset can be used for the minimum of the stratified sampling algorithm, at this time data from the sample survey sum is 4X, and average about 4 data from the sample survey determine
1 pixel, similarity reach 0.99 or more on display.Total evidence will appear the distortion phenomenon of Fig. 4 c when being divided into 1X subset,
Being divided into 2X, 4X, 6X, 8X or higher (being incremented by with even-multiple), similarity will gradually be incremented by, and similarity reaches 0.9999 left side when 50X
The right side, sampling pattern and full datagraphic are substantially completely consistent at this time.But totally more according to the subset number of distribution, sampling rate must
It is so slower.
In conclusion for the slow problem of current mass data plotting speed, according to drawing window width pixel number (X)
Total to obtain 4X data from the sample survey using stratified sampling algorithm according to 2X subset is divided into, this 4X data from the sample survey figure will at this time
It is totally consistent (0.99 or more similarity) according to pattern height.It is compared totally according to plotting speed, " single thread sampling+drawing " is whole
Body speed promotes about 70 times, and " thread pool sampling+drawing " bulk velocity promotes about 200 times.
High-efficiency layered sampling drawing practice and system of a kind of mass data provided in an embodiment of the present invention in time-domain, mesh
In preceding " the earth pulsation short-impending prediction real-time tracking analysis processing system " for being formally applied to author's research and development, the system is current
It, in next step will be to national ground during impending earthquake forecast applied to Gansu, Qinghai, Sichuan, Yunnan, Tibet, Xinjiang region is practiced
Shake system is promoted and applied.
In the above-described embodiments, it can all be realized by software.When using entirely or partly with computer program
The form of product realizes that the computer program product includes one or more computer instructions.It loads or holds on computers
When the row computer program instructions, entirely or partly generate according to process or function described in the embodiment of the present invention.It is described
Computer can be general purpose computer, special purpose computer, computer network or other programmable devices.The computer refers to
Order may be stored in a computer readable storage medium, or can to another computer from a computer readable storage medium
Storage medium transmission is read, for example, the computer instruction can be from a web-site, computer, server or data center
By it is wired (such as coaxial cable, optical fiber, Digital Subscriber Line (DSL) or wireless (such as infrared, wireless, microwave etc.) mode to
Another web-site, computer, server or data center are transmitted).The computer-readable storage medium can be with
It is any usable medium that can access of computer or includes the integrated server of one or more usable mediums, in data
The data storage devices such as the heart.The usable medium can be magnetic medium, (for example, floppy disk, hard disk, tape), optical medium (example
Such as, DVD) or semiconductor medium (such as solid state hard disk Solid State Disk (SSD)) etc..
The foregoing is merely illustrative of the preferred embodiments of the present invention, is not intended to limit the invention, all in essence of the invention
Made any modifications, equivalent replacements, and improvements etc., should all be included in the protection scope of the present invention within mind and principle.
Claims (5)
1. a kind of mass data time-domain high-efficiency layered sample drawing practice, which is characterized in that the mass data when
Between domain high-efficiency layered sampling drawing practice include:
Step 1 is sampled mass data using high-efficiency layered sampling algorithm according to drawing window width, obtains 4X pumping
Sample data；
Step 2 is drawn with 4X data from the sample survey, obtains full datagraphic.
2. mass data as described in claim 1 is in the high-efficiency layered sampling drawing practice of time-domain, which is characterized in that described
Mass data is specifically included in the high-efficiency layered methods of sampling of time-domain:
The first step, draw data sum=N, drawing window width=X pixel, data from the sample survey sum=4X；
Second step, N > 4X?
Third step, if it is not, unsample, draw data=total evidence；
4th step extracts 4X data from the sample survey if so, being sampled to total evidence；
5th step, draw data=4X data from the sample survey.
3. mass data as described in claim 1 is in the high-efficiency layered sampling drawing practice of time-domain, which is characterized in that described
The mass data high-efficiency layered methods of sampling is according to 2X subset is divided into, all data seek maximum respectively in each 2X subset totally
With minimum value and by the sequencing storage occurred, 4X data from the sample survey is obtained；It specifically includes:
Step 1: draw data sum=N, drawing window width=X pixel, data from the sample survey sum=4X；
Step 2:N > 4X?
Step 3: if it is not, unsample, draw data=total evidence；
Step 4: if so, class interval douJG=(double) data count N/ (2*X)；Every group of data amount check P=(rounding)
douJG；
Step 5: cycle calculations grouping serial number intLen [i]=(rounding) (douJG*i), wherein i=0 ... 2X+1, intLen [i]
Time interval (data count for representing each subset) value of each element of=0 ... N, intLen are as follows: P P ... P P+1P ... P
P+1P P increases 1 every several, remaining all P；
Step 6: cycle calculations sampled value floSam [i], i=0 ... 2X-1, floSam sum are 4X；
Step 7: extracting serial number all data between intLen [i] to intLen [i+1] -1, calculate maximum value floMax, most
Small value floMin, maximum value serial number intPosMax, minimum value serial number intPosMin；
Step 8: minimum value is formerly or maximum value is first, sequentially stores；
FloSam [2i]=intPosMin < intPosMax? floMin:floMax；
FloSam [2i+1]=intPosMin < intPosMax? floMax:floMin；
Does step 9: circulation terminate? if it is not, repeating step 6；
Step 10: if so, draw data=4X data from the sample survey.
4. a kind of realize mass data described in claims 1 to 3 any one in the high-efficiency layered sampling drawing practice of time-domain
Computer program.
5. a kind of realize mass data described in claims 1 to 3 any one in the high-efficiency layered sampling drawing practice of time-domain
Mass data time-domain high-efficiency layered sample drawing control system.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910101886.1A CN109920025B (en) | 2019-01-31 | 2019-01-31 | Efficient hierarchical sampling drawing method and system for mass data in time domain |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201910101886.1A CN109920025B (en) | 2019-01-31 | 2019-01-31 | Efficient hierarchical sampling drawing method and system for mass data in time domain |
Publications (2)
| Publication Number | Publication Date |
|---|---|
| CN109920025A true CN109920025A (en) | 2019-06-21 |
| CN109920025B CN109920025B (en) | 2023-10-27 |
Family
ID=66961128
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201910101886.1A Active CN109920025B (en) | 2019-01-31 | 2019-01-31 | Efficient hierarchical sampling drawing method and system for mass data in time domain |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN109920025B (en) |
Cited By (2)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114071269A (en) * | 2022-01-18 | 2022-02-18 | 武汉烽火信息集成技术有限公司 | Performance trend calculation method, medium, equipment and system of optical network access product |
| CN116069833A (en) * | 2022-12-30 | 2023-05-05 | 北京航天测控技术有限公司 | Mass data processing and visualizing method based on dynamic multi-level resolution self-adaptive analysis |
Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1754194A (en) * | 2002-12-30 | 2006-03-29 | 小利兰斯坦福大学理事会 | Methods and apparatus for interactive map-based analysis of digital video content |
| CN1878260A (en) * | 2006-07-14 | 2006-12-13 | 杭州国芯科技有限公司 | Multi-menu co-screen playing method |
| CN102262678A (en) * | 2011-08-16 | 2011-11-30 | 郑毅 | System for sampling mass data and managing sampled data |
| CN103077557A (en) * | 2013-02-07 | 2013-05-01 | 河北大学 | Adaptive hierarchical chest large data display implementation method |
| CN103955508A (en) * | 2014-04-29 | 2014-07-30 | 安徽中科大国祯信息科技有限责任公司 | Method for layering, storing and inquiring data of curve chart of project control system |
| CN103995684A (en) * | 2014-05-07 | 2014-08-20 | 广东粤铁瀚阳科技有限公司 | Method and system for synchronously processing and displaying mass images under ultrahigh resolution platform |
| CN105912562A (en) * | 2016-03-21 | 2016-08-31 | 华南理工大学 | Data visualization method and data visualization system based on hierarchical model |
| CN106599172A (en) * | 2016-12-10 | 2017-04-26 | 成都飞机工业（集团）有限责任公司 | Lossless fidelity visualization optimization method for mass data |
| CN106776733A (en) * | 2016-11-22 | 2017-05-31 | 北京航天自动控制研究所 | The data staging methods of sampling in data handling system |
| CN106802924A (en) * | 2016-12-19 | 2017-06-06 | 北京科胜永昌软件有限公司 | The cautious data of magnanimity observation system big gun based on multi-threaded parallel draw display methods |
-
2019
- 2019-01-31 CN CN201910101886.1A patent/CN109920025B/en active Active
Patent Citations (10)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN1754194A (en) * | 2002-12-30 | 2006-03-29 | 小利兰斯坦福大学理事会 | Methods and apparatus for interactive map-based analysis of digital video content |
| CN1878260A (en) * | 2006-07-14 | 2006-12-13 | 杭州国芯科技有限公司 | Multi-menu co-screen playing method |
| CN102262678A (en) * | 2011-08-16 | 2011-11-30 | 郑毅 | System for sampling mass data and managing sampled data |
| CN103077557A (en) * | 2013-02-07 | 2013-05-01 | 河北大学 | Adaptive hierarchical chest large data display implementation method |
| CN103955508A (en) * | 2014-04-29 | 2014-07-30 | 安徽中科大国祯信息科技有限责任公司 | Method for layering, storing and inquiring data of curve chart of project control system |
| CN103995684A (en) * | 2014-05-07 | 2014-08-20 | 广东粤铁瀚阳科技有限公司 | Method and system for synchronously processing and displaying mass images under ultrahigh resolution platform |
| CN105912562A (en) * | 2016-03-21 | 2016-08-31 | 华南理工大学 | Data visualization method and data visualization system based on hierarchical model |
| CN106776733A (en) * | 2016-11-22 | 2017-05-31 | 北京航天自动控制研究所 | The data staging methods of sampling in data handling system |
| CN106599172A (en) * | 2016-12-10 | 2017-04-26 | 成都飞机工业（集团）有限责任公司 | Lossless fidelity visualization optimization method for mass data |
| CN106802924A (en) * | 2016-12-19 | 2017-06-06 | 北京科胜永昌软件有限公司 | The cautious data of magnanimity observation system big gun based on multi-threaded parallel draw display methods |
Cited By (3)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN114071269A (en) * | 2022-01-18 | 2022-02-18 | 武汉烽火信息集成技术有限公司 | Performance trend calculation method, medium, equipment and system of optical network access product |
| CN116069833A (en) * | 2022-12-30 | 2023-05-05 | 北京航天测控技术有限公司 | Mass data processing and visualizing method based on dynamic multi-level resolution self-adaptive analysis |
| CN116069833B (en) * | 2022-12-30 | 2024-04-19 | 北京航天测控技术有限公司 | Mass data processing and visualizing method based on dynamic multi-level resolution self-adaptive analysis |
Also Published As
| Publication number | Publication date |
|---|---|
| CN109920025B (en) | 2023-10-27 |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| Hao et al. | Multi-resolution techniques for visual exploration of large time-series data | |
| Mihalisin et al. | Visualization and analysis of multi-variate data: a technique for all fields | |
| CN102737243B (en) | Method and device for acquiring descriptive information of multiple images and image matching method | |
| US8005992B1 (en) | Scalable storage and retrieval of multiple asynchronous signals | |
| CN109920025A (en) | A kind of mass data is sampled drawing practice and system in the high-efficiency layered of time-domain | |
| CN109725185A (en) | A kind of oscillograph and its operation method for realizing waveform fast Acquisition | |
| CN106570049B (en) | The curve QuickDraw method and its system of big data quantity in a kind of geological prospecting | |
| Parulek et al. | Implicit surfaces for interactive graph based cavity analysis of molecular simulations | |
| Yang et al. | A web based MDSplus data analysis and visualization system for EAST | |
| Peng et al. | Visual abstraction of dynamic network via improved multi-class blue noise sampling | |
| Lee et al. | Benchmarking community detection methods on social media data | |
| Tan et al. | MMSparse: 2D partitioning of sparse matrix based on mathematical morphology | |
| Siang et al. | An overview of immersive data visualisation methods using type by task taxonomy | |
| EP2888719A1 (en) | Method and system for data visualization | |
| Dasgupta et al. | Meta parallel coordinates for visualizing features in large, high-dimensional, time-varying data | |
| Yamasaki et al. | Reconstruction and simulation for three-dimensional morphological structure of insect neurons | |
| Wang et al. | Efficient Stratified Sampling Graphing Method for Mass Data | |
| CN112463844A (en) | Data processing method and device, electronic equipment and storage medium | |
| Baum | Introducing aesthetics to software visualization | |
| CN105069082A (en) | Plant fluorescence image sequence data processing method, Web server and system | |
| Williams et al. | SitaVis—Interactive situation awareness visualization of large datasets | |
| Alvanides et al. | Zone Design as a Spatial Analysis | |
| Hsieh et al. | Parallel tsunami simulation and visualization on tiled display wall using OpenGL Shading Language | |
| Becker et al. | HFM: Hierarchical Feature Moment Extraction for Multi-Omic Data Visualization | |
| Wang et al. | Application of Fisheye View in Temporal Spiral Visualization |
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 | ||
| GR01 | Patent grant | ||
| GR01 | Patent grant |