CN103729796A - Method and system for sample survey - Google Patents

Abstract

The embodiment of the invention discloses a method for a sample survey. The method includes the steps of selecting the range of samples and obtaining the service types of the samples and the sample number of the service types; according to the service types and the sample number, setting sampling survey cost and a sample distribution method, and estimating the total number of the samples and the sample quantity of the service types; selecting one service type, obtaining the sample quantity of the service type, and setting a sampling method to obtain the sample extraction quantity of the service; according to the sample distribution method and the sampling method, determining an overall calculation algorithm, and calculating an overall estimation mean value and an estimation variance of the sample quantity of the service according to the overall calculation algorithm, the sample quantity of the service and the sample extraction quantity of the service. The embodiment of the invention further discloses a system for the sample survey. According to the method and the system, the problems that in the prior art, the requirements for quality of professionals are high, and a sampling method is complex and low in efficiency, operated manually and not suitable for periodical sampling surveys are solved.

Description

A kind of method and system of sample survey

Technical field

The present invention relates to power system measuring technical field, relate in particular to a kind of method and system of sample survey.

Background technology

Sample survey has been widely used in the fields such as social economy, science and technology, nature, is that researchist obtains one of important means of statistics.Researchist carries out Sampling Survey by the decimation blocks of general statistical software conventionally.Traditional mode of operation is first business datum to be imported to statistical software, then samples by the decimation blocks of general statistical software, finally investigation result data importing is carried out to statistical inference to statistical software.

This mode has following shortcoming: statistical knowledge and the software programming requested knowledge of (1) which to concrete operations personnel is higher, and a code error of any one link all will cause infers mistake partially; (2) which cannot obtain many auxiliary variables that complex sampling relates in real time, and these variablees usually need real-time query Service Database just can obtain; (3) when which deal with data is recorded in business datums more than 1,000,000 ranks, efficiency is lower, needs manual operation and wastes a large amount of time for data transmission; (4) which is inefficient manual operation method, is not suitable for periodic sampling investigation business.

Summary of the invention

The object of the embodiment of the present invention is to provide a kind of method and system of sample survey, can solve that in prior art, professional personnel qualifications is high, sample mode is complicated and inefficient manual operation and be not suitable for the problem of periodic sampling investigation business.

In order to solve the problems of the technologies described above, the embodiment of the present invention provides a kind of method of sample survey, and described method comprises:

Select required sampling sample range, multiple types of service and the number of samples corresponding to each type of service of the sample that obtains sampling;

According to the described multiple types of service that obtain and number of samples corresponding to each type of service, sample survey cost and sample distribution method are set, estimate the sampling sample size of sampling sample total and each type of service;

Select the type of service of described sampling sample, obtain the sampling sample size of selected type of service, and the methods of sampling is set the sampling sample size of the described selected type of service obtaining is carried out to the extraction of sample, obtain the sampling amount of selected type of service;

According to sample distribution method and the methods of sampling of described setting, determine the overall algorithm of calculating, and according to described definite overall reckoning algorithm, described in the sampling sample size of selected type of service and the sampling amount of selected type of service that obtain, calculate overall estimation average and the estimation variance of the sampling sample size of selected type of service.

Wherein, described sampling sample range comprises the time of the sample of sampling, type of service and the sample survey department of sampling sample.

Wherein, described sample survey cost comprises the sample survey time of each type of service in sample survey number, the total number of days of sample survey and the each sample survey of each sample survey number of days department.

Wherein, the sampling sample size that estimates sampling sample total and each type of service described in is to obtain based on cost of sampling estimation or based on considering the Neyman layering estimation of cost.

Wherein, described sample distribution method comprise proportional allocation, do not consider cost Neyman layering distribution method, consider Neyman layering distribution method and the multistage distribution method of cost.

Wherein, the described methods of sampling comprises arbitrary sampling method and the PPS methods of sampling.

The embodiment of the present invention also provides a kind of system of sample survey, and described system comprises:

One Service Database interface module, contains for building the Service Database of sample range of sampling, and optional sampling sample is provided, and obtains multiple types of service and the number of samples corresponding to each type of service of described sampling sample;

One sampling sample size estimation block, for the multiple types of service and number of samples corresponding to each type of service that obtain described in basis, arranges sample survey cost and sample distribution method, estimates the sampling sample size of sampling sample total and each type of service;

One sampling plan setting module, for selecting the type of service of described sampling sample, obtain the sampling sample size of selected type of service, and the methods of sampling is set the sampling sample size of the described selected type of service obtaining is carried out to the extraction of sample, obtain the sampling amount of selected type of service;

One overall inference module, be used for according to sample distribution method and the methods of sampling of described setting, determine the overall algorithm of calculating, and according to described definite overall reckoning algorithm, described in the sampling sample size of selected type of service and the sampling amount of selected type of service that obtain, calculate overall estimation average and the estimation variance of the sampling sample size of selected type of service.

Wherein, described sample survey cost comprises the sample survey time of each type of service in sample survey number, the total number of days of sample survey and the each sample survey of each sample survey number of days department.

Wherein, the sampling sample size that estimates sampling sample total and each type of service described in is to obtain based on cost of sampling estimation or based on considering the Neyman layering estimation of cost.

Wherein, described sample distribution method comprise proportional allocation, do not consider cost Neyman layering distribution method, consider Neyman layering distribution method and the multistage distribution method of cost.

Implement the embodiment of the present invention, there is following beneficial effect:

1, with subscriber service system and Service Database seamless link, and according to customer service, need to set complex sampling algorithm and realize the robotization that statistical sampling is inferred, solve the insurmountable problem of general statistical software decimation blocks;

2,, aspect the selection of sampling sample, user can select the time of the sampling sample that needs investigation, type of service and the survey section of sampling sample in conjunction with practical business; In sampling distribution method and the methods of sampling, included the proportional allocation of current main flow, do not considered the Neyman(Niemann of cost) distribution method such as layering distribution method, the Neyman layering distribution method of considering cost and multistage distribution method, also comprise arbitrary sampling method, PPS(Probability Proportionate to Size Sampling, press scale proportional) methods of sampling such as the methods of sampling, and can select applicable sampling distribution method and the methods of sampling according to traffic performance; Aspect overall deduction, according to the methods of sampling, realize robotization, reduce manual operation ratio, raising efficiency;

3, at software architecture design aspect, the convenience of the operating personnel's that the framework based on B/S promotes greatly work, the data typing that is no matter sampling phase or investigation phase can be carried out whenever and wherever possible, can also use handheld terminal to carry out data typing and inquiry while investigating at the scene.

Accompanying drawing explanation

In order to be illustrated more clearly in the embodiment of the present invention or technical scheme of the prior art, to the accompanying drawing of required use in embodiment or description of the Prior Art be briefly described below, apparently, accompanying drawing in the following describes is only some embodiments of the present invention, for those of ordinary skills, do not paying under the prerequisite of creative work, the accompanying drawing that obtains other according to these accompanying drawings still belongs to category of the present invention.

The process flow diagram of the method for the sample survey that Fig. 1 provides for the embodiment of the present invention;

The structural representation of the sampling sample range that Fig. 2 provides for the embodiment of the present invention;

The structural representation of the sample distribution method that Fig. 3 provides for the embodiment of the present invention and methods of sampling combination;

The structural representation of the system of the sample survey that Fig. 4 embodiment of the present invention provides.

Embodiment

In order to make object of the present invention, technical scheme and advantage clearer, below in conjunction with drawings and Examples, the present invention is further elaborated.Should be appreciated that specific embodiment described herein, only in order to explain the present invention, is not intended to limit the present invention.

As shown in Figure 1, in the embodiment of the present invention, propose a kind of method of sample survey, described method comprises:

Step S101, select required sampling sample range, multiple types of service and the number of samples corresponding to each type of service of the sample that obtains sampling;

As shown in Figure 2, sampling sample range comprises the time of the sample of sampling, type of service and the sample survey department of sampling sample, business personnel can select required sampling sample (in as Fig. 2, " √ " is selected) in sampling sample range, obtain multiple sampling samples, this sampling sample comprises multiple types of service, and each type of service is to there being different numbers of samples.For example: select A1 institute, in Dec, the 2013 business inspection of A3 institute, obtain 2 sampling samples, in A1, A3 institute, there is 2 types of service (i, j), wherein, in i kind business, A1 institute is to there being 300,000 numbers of samples, A3 institute is to there being 500,000 numbers of samples, and in j kind business, A1 institute is to there being 1,000,000 numbers of samples, and A3 institute is to there being 400,000 numbers of samples.

The multiple types of service and number of samples corresponding to each type of service that described in step S102, basis, obtain, arrange sample survey cost and sample distribution method, estimates the sampling sample size of sampling sample total and each type of service;

Sample survey cost comprises the sample survey time of each type of service in sample survey number, the total number of days of sample survey and the each sample survey of each sample survey number of days department;

For example: in A1 institute, in in Dec, the 2013 business inspection of A3 institute, the sample survey cost arranging is sample survey number 2 people, the total number of days of sample survey 20 days, the first day A1 5 minutes kind business survey time of i, 30 minutes j kind business survey time, the A3 15 minutes kind business survey time of i, 20 minutes j kind business survey time, second day A1 5 minutes kind business survey time of i, 10 minutes j kind business survey time, the A3 25 minutes kind business survey time of i, 20 minutes j kind business survey time etc., the like carry out the setting of sample survey cost.

Sample distribution method comprises proportional allocation, does not consider the Neyman layering distribution method of cost, considers Neyman layering distribution method and the multistage distribution method of cost.

To sum up, business personnel is according to multiple types of service of the sampling sample obtaining and number of samples corresponding to each type of service, as required or be fixedly installed sample survey cost and sample distribution method, estimate the sampling sample size of sampling sample total and each type of service.

In the embodiment of the present invention, provide two kinds of methods that estimate the sampling sample size of sampling sample total and each type of service: one, based on cost of sampling evaluation method; Two, the Neyman layering evaluation method based on considering cost.

The first based on cost of sampling evaluation method in, (for example add up the number of samples total amount of certain type service (i kind business), dropping into 2 staff every month investigates, investigation number of days is 20 days, number of samples in each sample survey every day department in the different sample survey time: 5000 of the numbers of samples that 5 minutes first day A1 institute's time obtained, 12000 of the numbers of samples that 15 minutes A3 institute's time obtained, 3000 of the numbers of samples that second day 5 minutes A1 institute's time obtained, 18000 of numbers of samples that 25 minutes A3 institute's time obtained etc., until the summation of A1 institute and A3 institute number of samples in arrange 20 days), be designated as T _i, mean sample control time (for example, A1 investigate the mean value of T.T. with A3 institute in these 20 days), be designated as t, the sample size of sampling n _ifor

sampling sample total n is

At the second, based on considering, in the Neyman layering evaluation method of cost, suppose that expense meets relational expression

wherein c ₀for basic expense, c _hbe the cost that under this business, h layer is carried out unit sampling, C is total cost.Expense, as input parameter, directly has influence on the sample size of sampling, the sampling sample size n of Neyman to certain type service (i kind business) _icomputing formula as follows:

sampling sample total n is

wherein w _hbe the overall layer power of h layer, s _hit is the population variance of h layer.

Step S203, select the type of service of described sampling sample, obtain the sampling sample size of selected type of service, and the methods of sampling is set the sampling sample size of the described selected type of service obtaining is carried out to the extraction of sample, obtain the sampling amount of selected type of service;

The methods of sampling comprises arbitrary sampling method and the PPS methods of sampling; As shown in Figure 3, sample distribution method and the methods of sampling can respectively select one to combine.

Business personnel is by the type of service of selective sampling sample, obtain the sampling sample size of selected type of service, and the methods of sampling of this type of service is set, and the sampling sample size of this type of service obtaining is carried out to the extraction of sample, obtain the sampling amount of selected type of service; For example: select i kind business, just can determine the sampling sample size n of i kind business _i, simultaneously according to this type of service, it is the PPS methods of sampling that the methods of sampling is set, and obtains the sampling amount n of i kind business _i,h.

In the embodiment of the present invention, provide two kinds of methods of samplings: arbitrary sampling method and the PPS methods of sampling, wherein, adopt La Xili method implementation in the PPS methods of sampling.

Take the PPS methods of sampling as example, need to determine in advance an auxiliary variable M _ias unit tolerance, (be simply designated as M _i).For example, for power supply administration's different business, the auxiliary variable of system is as described in Table 1:

Table 1:

The algorithm design of PPS sampling La Xili method is as follows:

Input: all samples of layer, tolerance M _ioutput: the sampling amount (n of layer _i,hindividual sample);

The first step: tolerance is carried out to necessary conversion, make it become positive number, such as being multiplied by 10 or 100 etc.

Second step: take out maximum tolerance, be designated as M, wherein,

n _ifor the sampling sample size of this layer;

The 3rd step: simultaneously to interval [1, M] and [1, N _i] generate respectively a random number, be designated as respectively m and i, if M _i>=m, i sample drawn;

The 4th step: repeat the 3rd step, until obtain the sampling amount n of regulation in advance _i,hafter, stop sampling.

Step S204, according to sample distribution method and the methods of sampling of described setting, determine the overall algorithm of calculating, and according to described definite overall reckoning algorithm, described in the sampling sample size of selected type of service and the sampling amount of selected type of service that obtain, calculate overall estimation average and the estimation variance of the sampling sample size of selected type of service;

According to the various combination of sample distribution method and the methods of sampling, obtain different overall reckoning algorithms, according to the overall reckoning algorithm obtaining, the sampling sample size of selected type of service obtaining and the sampling amount of selected type of service, calculate overall estimation average and the estimation variance of the sampling sample size of selected type of service; Wherein, totally calculate that algorithm comprises that stratified random smapling population mean reckoning algorithm, layering unequal probability (PPS) sample population ratio are calculated, multiphase sampling population mean is estimated scheduling algorithm.

For example, (1) first method, stratified random smapling population mean are calculated:

Population mean $\stackrel{\‾}{Y}=\frac{1}{N}\underset{h=1}{\overset{L}{\mathrm{\Σ}}}\underset{i=1}{\overset{N_h}{\mathrm{\Σ}}}{Y}_{h,i}=\underset{h=1}{\overset{L}{\mathrm{\Σ}}}{W}_h{\stackrel{\‾}{Y}}_h;$

Overall estimation average is ${\stackrel{=}{y}}_{\mathrm{st}}=\underset{h=1}{\overset{L}{\mathrm{\Σ}}}{W}_h{\stackrel{\‾}{y}}_h=\frac{1}{N}\underset{i=1}{\overset{L}{\mathrm{\Σ}}}{N}_h{\stackrel{\‾}{y}}_h;$

The variance of this estimator is $v\left({\stackrel{=}{y}}_{\mathrm{st}}\right)=\underset{h=1}{\overset{L}{\mathrm{\Σ}}}\frac{W_h^2{S}_h^2}{n_h}-\frac{\underset{h=1}{\overset{L}{\mathrm{\Σ}}}{W}_h{S}_h^2}{N};$

$v\left({\stackrel{=}{y}}_{\mathrm{st}}\right)=\underset{h=1}{\overset{L}{\mathrm{\Σ}}}\frac{W_h^2{s}_h^2}{n_h}(1-f_h)$

This estimation variance is $=\underset{h=1}{\overset{L}{\mathrm{\Σ}}}\frac{W_h^2{s}_h^2}{n_h}-\frac{\underset{h=1}{\overset{L}{\mathrm{\Σ}}}{W}_h{s}_h^2}{N};$ Wherein,

(2) second method, layering unequal probability (PPS) overall rate are calculated:

Sampling sample size is n, and i layer number of samples is N _i; Wherein, sampling i unit of sample is y _i, i=1,2 ..., n, the auxiliary variable of i unit is M _i, i=1,2 ..., N _i;

Obtain the total value M of auxiliary variable ₀for

i the probability z that unit is drawn _ifor

wherein, $\underset{i=1}{\overset{N}{\mathrm{\Σ}}}{z}_i=1;$

Order $q_i=\frac{y_i}{{\mathrm{Nz}}_i},i=\mathrm{1,2}...,n;$

Overall estimation average is estimated as q exactly

$s^2\left(\stackrel{\‾}{q}\right)=\frac{1}{n(n-1)}\underset{i=1}{\overset{n}{\mathrm{\Σ}}}{(p_i-\stackrel{\‾}{q})}^2$

estimation variance, population variance is $=\frac{1}{n(n-1)}(\underset{i=1}{\overset{n}{\mathrm{\Σ}}}{p}_i^2-{n\stackrel{\‾}{q}}^2)\·$

(3) the third method, multiphase sampling population mean are estimated:

The estimation of multiphase sampling divides three steps: average and the variance of 1, estimating secondary unit; 2, estimate average and the variance of primary unit; 3, estimate overall average and variance.

Multiphase sampling need to do to unit at different levels the estimation of average and variance, therefore by the mode of recursion, obtains.Wherein, primary unit number is K, and the secondary unit number of i primary unit is M _i, three grades of unit numbers of j secondary unit of i primary unit are N _ij, j secondary unit l of i primary unit three grades of unit are $Y_{\mathrm{ijl}}\left\{\begin{array}{c}i=\mathrm{1,2},...,K\\ j=\mathrm{1,2},...,M_i\\ l=\mathrm{1,2},...,N_{\mathrm{ij}}\end{array}\right\},$ The number that contains three grades of unit in i primary unit is

total number containing three grades of unit in overall is

overall total amount is $\tilde{Y}=\underset{i=1}{\overset{K}{\mathrm{\Σ}}}\underset{j=1}{\overset{M_i}{\mathrm{\Σ}}}\underset{l}{\overset{N_{\mathrm{ij}}}{\mathrm{\Σ}}}{Y}_{\mathrm{ijl}},$ Population mean is $\stackrel{\‾}{Y}=\frac{1}{N}\tilde{Y}.$

What three stages were concrete is estimated as follows:

The first step, secondary unit are estimated: to each secondary unit, its Estimation of Mean is

to each secondary unit, its summation is

the variance of average is estimated as

wherein, $s_{\mathrm{ij}}^2=\frac{1}{n_{\mathrm{ij}}-1}\underset{l=1}{\overset{n_{\mathrm{ij}}}{\mathrm{\Σ}}}{(y_{\mathrm{ijl}}-{\stackrel{\‾}{y}}_{\mathrm{ij}})}^2,f_{\mathrm{ij}}=\frac{n_{\mathrm{ij}}}{N_{\mathrm{ij}}};$ The variance of summation is estimated as $s_{{\tilde{y}}_{\mathrm{ij}}}^2=N_{\mathrm{ij}}^2{s}_{{\stackrel{\‾}{y}}_{\mathrm{ij}}}^2.$

Second step, primary unit are estimated: to each primary unit, its Estimation of Mean is

to each primary unit, its summation is estimated as

the variance of summation is estimated as $s_{{\tilde{y}}_i}^2=\frac{M_i^2}{m_i}{\hat{V}}_{1i}+\frac{M_i^2}{m_i}(1-f_i){\hat{V}}_{2i};$ Wherein, ${\hat{V}}_{1i}=\frac{1}{m_i}\underset{j=1}{\overset{m_i}{\mathrm{\Σ}}}{s}_{{\tilde{y}}_{\mathrm{ij}}}^2,$ ${\hat{V}}_{2i}=\frac{1}{m_i-1}\underset{j=1}{\overset{m_i}{\mathrm{\Σ}}}{({\tilde{y}}_{\mathrm{ij}}-\frac{1}{m_i}\underset{j=1}{\overset{m_i}{\mathrm{\Σ}}}{\tilde{y}}_{\mathrm{ij}})}^2-{\hat{V}}_{1i},$ $f_i=\frac{m_i}{M_i};$ The variance of average is estimated as $s_{{\stackrel{\‾}{y}}_i}^2=\frac{1}{M_i^2}{s}_{{\tilde{y}}_i}^2.$

The estimation of the 3rd step, population mean and variance: totally estimate that summation is

overall estimation average is $\stackrel{\‾}{y}=\frac{1}{N}\tilde{y}=\frac{1}{N}\frac{K}{k}\underset{i=1}{\overset{k}{\mathrm{\Σ}}}{\tilde{y}}_i;$

Overall summation estimation variance is $s_{\tilde{y}}^2=\frac{K^2}{k}{\hat{V}}_1+\frac{K^2}{k}(1-f){\hat{V}}_2;$ Wherein, ${\hat{V}}_1=\frac{1}{k}\underset{i=1}{\overset{k}{\mathrm{\Σ}}}{s}_{{\tilde{y}}_i}^2,$ ${\hat{V}}_2=\frac{1}{k-1}\underset{i=1}{\overset{k}{\mathrm{\Σ}}}{({\tilde{y}}_i-\frac{1}{k}\underset{i=1}{\overset{k}{\mathrm{\Σ}}}{\tilde{y}}_i)}^2-{\hat{V}}_1,f=\frac{k}{K};$ Population mean estimation variance is $s_{\stackrel{\‾}{y}}^2=\frac{1}{N^2}{s}_{\tilde{y}}^2.$

Implement the embodiment of the present invention, there is following beneficial effect:

1, with subscriber service system and Service Database seamless link, and according to customer service, need to set complex sampling algorithm and realize the robotization that statistical sampling is inferred, solve the insurmountable problem of general statistical software decimation blocks;

2,, aspect the selection of sampling sample, user can select the time of the sampling sample that needs investigation, type of service and the survey section of sampling sample in conjunction with practical business; In sampling distribution method and the methods of sampling, include the proportional allocation of current main flow, do not considered the distribution methods such as the Neyman layering distribution method of cost, the Neyman layering distribution method of considering cost and multistage distribution method, also comprise the methods of samplings such as arbitrary sampling method, the PPS methods of sampling, and can select applicable sampling distribution method and the methods of sampling according to traffic performance; Aspect overall deduction, according to the methods of sampling, realize robotization, reduce manual operation ratio, raising efficiency;

3, at software architecture design aspect, the convenience of the operating personnel's that the framework based on B/S promotes greatly work, the data typing that is no matter sampling phase or investigation phase can be carried out whenever and wherever possible, can also use handheld terminal to carry out data typing and inquiry while investigating at the scene.

As shown in Figure 4, the embodiment of the present invention also provides a kind of system of sample survey, and described system comprises:

One Service Database interface module 410, contains for building the Service Database of sample range of sampling, and optional sampling sample is provided, and obtains multiple types of service and the number of samples corresponding to each type of service of described sampling sample;

One sampling sample size estimation block 420, for the multiple types of service and number of samples corresponding to each type of service that obtain described in basis, sample survey cost and sample distribution method are set, estimate the sampling sample size of sampling sample total and each type of service;

One sampling plan setting module 430, for selecting the type of service of described sampling sample, obtain the sampling sample size of selected type of service, and the methods of sampling is set the sampling sample size of the described selected type of service obtaining is carried out to the extraction of sample, obtain the sampling amount of selected type of service;

One overall inference module 440, be used for according to sample distribution method and the methods of sampling of described setting, determine the overall algorithm of calculating, and according to described definite overall reckoning algorithm, described in the sampling sample size of selected type of service and the sampling amount of selected type of service that obtain, calculate overall estimation average and the estimation variance of the sampling sample size of selected type of service.

Wherein, described sampling sample range comprises the time of the sample of sampling, type of service and the sample survey department of sampling sample.

Wherein, described sample survey cost comprises the sample survey time of each type of service in sample survey number, the total number of days of sample survey and the each sample survey of each sample survey number of days department.

Wherein, the sampling sample size that estimates sampling sample total and each type of service described in is to obtain based on cost of sampling estimation or based on considering the Neyman layering estimation of cost.

Wherein, described sample distribution method comprise proportional allocation, do not consider cost Neyman layering distribution method, consider Neyman layering distribution method and the multistage distribution method of cost.

Wherein, the described methods of sampling comprises arbitrary sampling method and the PPS methods of sampling.

In embodiments of the present invention, the system of sample survey by setting up Service Database and providing corresponding data-interface and operation system seamless link in Service Database interface module 410, and provide optional sampling sample, multiple types of service and the number of samples corresponding to each type of service of sample obtain sampling, multiple types of service and number of samples corresponding to each type of service that in sampling sample size estimation block 420, basis obtains, sample survey cost and sample distribution method are set, estimate the sampling sample size of sampling sample total and each type of service, the type of service of selective sampling sample in sampling plan setting module 430, obtain the sampling sample size of selected type of service, and the methods of sampling is set the sampling sample size of the selected type of service obtaining is carried out to the extraction of sample, obtain the sampling amount of selected type of service, in overall inference module 440 according to the sample distribution method and the methods of sampling that arrange, determine the overall algorithm of calculating, and according to the overall reckoning algorithm of determining, the sampling sample size of the selected type of service obtaining and the sampling amount of selected type of service, calculate overall estimation average and the estimation variance of the sampling sample size of selected type of service, thereby in solution prior art, professional personnel qualifications is high, complicated and the inefficient manual operation of sample mode and be not suitable for the problem of periodic sampling investigation business

One of ordinary skill in the art will appreciate that all or part of step realizing in above-described embodiment method is can carry out the hardware that instruction is relevant by program to complete, described program can be stored in a computer read/write memory medium, described storage medium, as ROM/RAM, disk, CD etc.

The foregoing is only preferred embodiment of the present invention, not in order to limit the present invention, all any modifications of doing within the spirit and principles in the present invention, be equal to and replace and improvement etc., within all should being included in protection scope of the present invention.

Claims (10)

1. a method for sample survey, is characterized in that, described method comprises:

Select required sampling sample range, multiple types of service and the number of samples corresponding to each type of service of the sample that obtains sampling;

According to the described multiple types of service that obtain and number of samples corresponding to each type of service, sample survey cost and sample distribution method are set, estimate the sampling sample size of sampling sample total and each type of service;

Select the type of service of described sampling sample, obtain the sampling sample size of selected type of service, and the methods of sampling is set the sampling sample size of the described selected type of service obtaining is carried out to the extraction of sample, obtain the sampling amount of selected type of service;

According to sample distribution method and the methods of sampling of described setting, determine the overall algorithm of calculating, and according to described definite overall reckoning algorithm, described in the sampling sample size of selected type of service and the sampling amount of selected type of service that obtain, calculate overall estimation average and the estimation variance of the sampling sample size of selected type of service.

2. the method for claim 1, is characterized in that, described sampling sample range comprises the time of the sample of sampling, type of service and the sample survey department of sampling sample.

3. the method for claim 1, is characterized in that, described sample survey cost comprises the sample survey time of each type of service in sample survey number, the total number of days of sample survey and the each sample survey of each sample survey number of days department.

4. the method for claim 1, is characterized in that, the sampling sample size of described sampling sample total and each type of service is to obtain based on cost of sampling estimation or based on considering the Neyman layering estimation of cost.

5. the method for claim 1, is characterized in that, described sample distribution method comprises proportional allocation, do not consider the Neyman layering distribution method of cost, considers Neyman layering distribution method and the multistage distribution method of cost.

6. the method for claim 1, is characterized in that, the described methods of sampling comprises arbitrary sampling method and the PPS methods of sampling.

7. a system for sample survey, is characterized in that, described system comprises:

One Service Database interface module, contains for building the Service Database of sample range of sampling, and optional sampling sample is provided, and obtains multiple types of service and the number of samples corresponding to each type of service of described sampling sample;

One sampling sample size estimation block, for the multiple types of service and number of samples corresponding to each type of service that obtain described in basis, arranges sample survey cost and sample distribution method, estimates the sampling sample size of sampling sample total and each type of service;

One sampling plan setting module, for selecting the type of service of described sampling sample, obtain the sampling sample size of selected type of service, and the methods of sampling is set the sampling sample size of the described selected type of service obtaining is carried out to the extraction of sample, obtain the sampling amount of selected type of service;

One overall inference module, be used for according to sample distribution method and the methods of sampling of described setting, determine the overall algorithm of calculating, and according to described definite overall reckoning algorithm, described in the sampling sample size of selected type of service and the sampling amount of selected type of service that obtain, calculate overall estimation average and the estimation variance of the sampling sample size of selected type of service.

8. system as claimed in claim 7, is characterized in that, described sample survey cost comprises the sample survey time of each type of service in sample survey number, the total number of days of sample survey and the each sample survey of each sample survey number of days department.

9. system as claimed in claim 7, is characterized in that, the sampling sample size of described sampling sample total and each type of service is to obtain based on cost of sampling estimation or based on considering the Neyman layering estimation of cost.

10. system as claimed in claim 7, is characterized in that, described sample distribution method comprises proportional allocation, do not consider the Neyman layering distribution method of cost, considers Neyman layering distribution method and the multistage distribution method of cost.

Images