CN108470033A - A kind of city public bicycle system borrows the visual analysis method of also pattern - Google Patents
A kind of city public bicycle system borrows the visual analysis method of also pattern Download PDFInfo
- Publication number
- CN108470033A CN108470033A CN201810101548.3A CN201810101548A CN108470033A CN 108470033 A CN108470033 A CN 108470033A CN 201810101548 A CN201810101548 A CN 201810101548A CN 108470033 A CN108470033 A CN 108470033A
- Authority
- CN
- China
- Prior art keywords
- user
- website
- riding
- public bicycles
- 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.)
- Pending
Links
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
- G06F16/904—Browsing; Visualisation therefor
-
- G—PHYSICS
- G06—COMPUTING; CALCULATING OR COUNTING
- G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
- G06Q30/00—Commerce
- G06Q30/02—Marketing; Price estimation or determination; Fundraising
- G06Q30/0201—Market modelling; Market analysis; Collecting market data
-
- G—PHYSICS
- G07—CHECKING-DEVICES
- G07F—COIN-FREED OR LIKE APPARATUS
- G07F17/00—Coin-freed apparatus for hiring articles; Coin-freed facilities or services
- G07F17/0042—Coin-freed apparatus for hiring articles; Coin-freed facilities or services for hiring of objects
- G07F17/0057—Coin-freed apparatus for hiring articles; Coin-freed facilities or services for hiring of objects for the hiring or rent of vehicles, e.g. cars, bicycles or wheelchairs
Landscapes
- Engineering & Computer Science (AREA)
- Business, Economics & Management (AREA)
- Finance (AREA)
- Strategic Management (AREA)
- Accounting & Taxation (AREA)
- Theoretical Computer Science (AREA)
- Development Economics (AREA)
- General Physics & Mathematics (AREA)
- Physics & Mathematics (AREA)
- Entrepreneurship & Innovation (AREA)
- Data Mining & Analysis (AREA)
- Databases & Information Systems (AREA)
- Game Theory and Decision Science (AREA)
- Economics (AREA)
- Marketing (AREA)
- General Business, Economics & Management (AREA)
- General Engineering & Computer Science (AREA)
- Management, Administration, Business Operations System, And Electronic Commerce (AREA)
Abstract
The invention discloses the visual analysis method that a kind of city public bicycle system borrows also pattern, the present invention first collects public bicycles system data and is pre-processed to it；Based on user perspective, designing user constitutes analysis chart, it shows the composition of user type, is based on time visual angle, variation diagram is ridden in designing user type and the annual amount of the riding variation spiral figure of duration relational graph of riding, design, design day and the user based on calendar figure rides ratio chart；Based on spatial view, design station space of points distribution map designs ride range distribution figure and design website clustering figure；More properties views are designed, analysis borrows vehicle the influence of also quantity under different weather situation, the effect of date property factor.The present invention can effectively improve understanding of the public bicycles system manager to system operation situation, improve the ability of public bicycles data law discovery, and the decision for station field signal, vehicle scheduling provides foundation.
Description
Technical field
The invention belongs to information technology fields, and the visualization of also pattern is borrowed more particularly to a kind of city public bicycle system
Analysis method.
Background technology
Public bicycles system is a kind of shared Cycle Hire service, is that a kind of novel, green public transport goes out
Row pattern.Public bicycles system is the component part of public transport, is the auxiliary and supplement of urban mass-transit system, is conducive to solve
Certainly urban transportation " last one kilometer " problem is of great significance to alleviating traffic pressure, realization energy-saving and emission-reduction etc..Nearly 10 years
Come, the lot of domestic and international city such as Shanghai, Hangzhou, Chicago, New York, Washington all establishes public bicycles system.
In the city for possessing public bicycles system, public bicycles website is typically distributed across traffic route both sides, often
A website includes multiple parking stake positions for being used for parking public bicycles.Public bicycles user generally comprises casual user and note
Volume user registers user's majority as working clan, and casual user is usually the lower user of public bicycles frequency that rides.User holds
IC card public bicycles website near departure place leases public bicycles by swiping the card, the public bicycles near destination
Operation of returning the car is carried out on the empty stake position of website.
User rents/goes back the data of public bicycles generation, such as every time when using the system：It leases the ID of bicycle, rent
Vehicle time, Site ID of hiring a car, site name of hiring a car, time of returning the car, Site ID of returning the car, site name of returning the car etc. are all passed by network
It is defeated to arrive data center.Public bicycles administrative staff do not have the data analysis capabilities of profession, are unable to fully rent/go back using user
" digital footprint " left by vehicle finds the rule of riding of public bicycles system user, to system carry out effectively management and
Decision.The characteristics of according to public bicycles system data, carries out visual analysis, Neng Goubang using effective visual analysis method
Public bicycles system manager is helped to find to contain the rule in public bicycles data.But existing public bicycles
Visual analysis method comes the user's composition and time-space attribute of display data, display effect list using block diagram, pie chart, line chart etc.
One and do not fully consider that influencing public bicycles rides the other factors (such as weather, festivals or holidays) of quantity.
Therefore, it is necessary to binding time, weather, dates etc. because usually designing visualization view, allow public bicycles administrator
Member can intuitively have found the rule of the rent of user/also by the visual pattern of design, and so as to help, they preferably manage
Public bicycles system preferably carries out the scheduling of public bicycles, alleviates the pressure of urban transportation.
Invention content
In view of the deficiencies of the prior art, the present invention proposes the visualizations that a kind of city public bicycle system borrows also pattern
Analysis method.
A kind of city public bicycle system borrows the visual analysis method of also pattern, this method to specifically include following step
Suddenly：
Step 1：It collects public bicycles system data and it is pre-processed.
Step 2：Based on user perspective, designing user constitutes analysis chart, shows the composition of user type, while in same figure
Gender and the composition at age are also provided in shape.
Step 3：Based on time visual angle, designing user type and duration relational graph of riding；It designs the whole year amount of riding and changes spiral shell
Rotation figure, shows the changing rule of public bicycles user's whole year amount of riding；Design day is ridden variation diagram, shows different user types
Daily changing rule of riding；It devises the user based on calendar figure to ride ratio chart, displaying different user types are daily to ride
Capable ratio.
Step 4：Based on spatial view, design station space of points distribution map shows the spatial distribution of city public bicycle；If
Meter is ridden range distribution figure, and displaying different user types are ridden the distribution situation of distance；Website clustering figure is designed, to public
The cluster result of bicycle website carries out visual analysis；
Step 5：More properties views are designed, analysis borrows also vehicle under different weather situation, the effect of date property factor
The influence of quantity.
It rides preferably, the collection public bicycles system data includes website essential information and user and records letter
Breath；Wherein website essential information includes Site ID, site name, site address, latitude, longitude, website on-line time and website
Stake digit；User record information of riding includes bicycle ID, card number, Site ID of hiring a car, time of chartering, can moor when hiring a car
Vehicle field, Site ID of returning the car, time of returning the car, the when of returning the car can park stake position, the duration of riding, user type, user's gender and
Year of birth, wherein user type include logical card user and year card user, and the record of riding of middle aged card user includes gender
And the age, it does not include gender and age to lead in the record of riding of card user；The pretreatment is：Deleting website of partly returning the car is
The record of riding of null value, deleting duplicated data and the data other than effective time, delete year card user ride record in gender or
Age is the data of null value, is ridden distance according to the calculation of longitude ＆ latitude of starting and purpose website, is provided using Baidu map API
Translate methods realize conversion of the GPSS84 coordinate systems to Baidu's coordinate system.
Preferably, step 2 specifically, using Visualization Framework Echarts interface, ride according to public bicycles
The statistical analysis of record will block male and female in ride number and the proportionate relationship of member and Tong Ka member, year card member year
Ride number and the proportionate relationship of user, different age group ride number and proportionate relationship is visualized with circular chart；It is right
Ride number and the proportion for answering the sector region of circular chart to include the certain customers.
Preferably, step 3 is specifically, user type and ride duration relational graph of the design based on stacking bar chart；Often
A bar shaped represents the number of riding of user in some duration of riding, wherein number of riding includes blocking member and Tong Ka member, year in year
Two different color bar shapeds are respectively adopted to indicate that the number of riding of male and female user, logical card user are adopted again in card member
It is indicated with the third color other than above two color；The height that each user type is stacked is higher, represents the time
The amount of riding of user is bigger in section, and vice versa；
The wherein annual amount of riding variation spiral figure is variation diagram of riding helical structure whole year of the design based on D3.js, by one
Year be unfolded come spiral using 12 kinds of different colors in 12 month, the quantity of riding on the date is checked in each bar shaped.
It rides variation diagram in day of the design based on area-graph；Wherein different user types use and user type and duration of riding
Relational graph identical Color scheme indicates；Abscissa indicates that period, ordinate indicate the difference of the period in area-graph
The data of riding of user type, the height of broken line is higher, indicates that date user's amount of riding is bigger；User is selected by the period
Component selectes special time period, and the data of the period will amplify；
User based on calendar figure rides ratio chart, shows time and the relationship that user is constituted；When user selects some moon
When part, which is precipitated the amount of riding that each user type is daily in this month, and is plotted to the moon under a proportional relationship
In the corresponding calendar of part；The data of different user types, the bigger shared fan section of data volume are mapped using different colors
Domain is bigger.
Preferably, the website spatial distribution map, which is the basic letter based on public bicycles system website
Breath inquiry shows the dense degree of website distribution in certain region according to scaling, by the intensive journey of website using polymerization technique
Degree is mapped as the size of radius；The polymerization radius in the more then region of certain region website is bigger, and the website in certain region is relatively few then should
The polymerization radius in region is smaller；
The range distribution figure of riding, data statistics of the view based on the entire public bicycles system amount of riding point
Analysis shows the quantity of riding of certain in section different user types of riding using bar chart is stacked；
The design website clustering figure；Using Unsupervised clustering algorithm：AP clustering algorithms are for public bicycles
Website carries out clustering；
First build similar matrix：
Assuming that there is n website in public bicycles site table, m region is finally required division into；S=S1, S2,
S...Sn } indicate public bicycles Website Hosting；A={ A1, A2, A3...Am } indicates to need the cluster that divides；dijIt indicates
Space length between any two website i and j, rijIndicate that the public bicycles between website i and j are ridden recorded amounts, i.e., from
Website i hires a car and returns the car and leased in website j and the public bicycles given back in website i record summation, wherein i, j in website j
∈S；Space length matrix D is is established respectively to all websites of public bicycles and public bicycles ride and record matrix Rel；
Dis is distance matrix, and setting means is as shown in formula 1.1, wherein dijIndicate website i to the space of website j away from
From.
Rel is record matrix of riding, and set-up mode is as shown in formula 1.2.Wherein rijIndicate the public affairs between website i and website j
Cycling recorded amounts altogether.
The model by public bicycles ride record relational matrix and space length matrix be combined, between structure the website
Similar matrix, in this, as the foundation of cluster；Before the record relationship that will ride is attached to distance matrix as weight, logarithm is needed
According to being normalized.
Data are normalized using such as 1.3 mode of formula, wherein rmaxIndicate the maximum value in Rel matrixes,
rminIndicate the minimum value in Rel matrixes；By formula it is found that if rent/also relationship between two websites is stronger, wijValue get over
It is small.
By the value w after normalizationijWith website distance value dijThe similar matrix Sn of website can be obtained by blending：
Then the topological structure of entire website is abstracted as graph structure；Three centrality of figure are calculated, and constitute matrix：
A variable α is finally introducing by two similar matrix SnAnd StIt is integrated into new similar matrix S i.e. S=α Sn+(1-
α)St；
It is ridden through the above steps according to site location and user and records structure similar matrix, then passed through AP and cluster calculation
Method obtains public bicycles website cluster result；
Website clustering figure is finally drawn according to cluster result；Pair of map is mapped to according to the latitude and longitude information of website
Different colors is answered on position and is mapped as the website of different clusters, the mark color of the same cluster website is identical.
Preferably, the more properties views of the design specifically include, generates influence factor tables of data, storage more than one and give
The date property and Weather property of every day in the setting analysis period；Wherein Weather property has seven attribute values：Fine day, it is cloudy,
Shower, moderate rain, heavy rain, snows at light rain；Wherein date property includes：Working day, weekend, festivals or holidays；Not using thermodynamic chart displaying
It rides distribution situation with the user under weather and date, user's amount of riding in certain region is bigger, and color is deeper, and vice versa；
Design more properties views：It is time shaft assembly below the figure, user selects data set by using time shaft assembly
Date in time range, figure top design weather component, the date selected according to time shaft assembly are inquired weather feelings
Condition, intermediate region are thermodynamic chart, show that the heating power that user rides is distributed according to the date of time shaft selection；
Design multi-attribute analysis figure：The matrix arranged using two rows four is displayed side by side 8 not under same date and weather conditions
Heating power situation；First row is the tourist season, and secondary series indicates that off-season, the first row indicate that working day fine day, secondary series indicate
Working day moderate rain, third row indicate that weekend fine day, the 4th row indicate rainy day at weekend or snowy day.
Advantageous effect：The characteristic of the method for the present invention and innovation be, by generalized time, spatial view, for it is public from
The characteristics of travelling data, designs visual representation method, contains the rule in public bicycles data by visual diagrammatic representation
Rule；A kind of new visual representation method " helical structure whole year ride variation diagram " is proposed, passes through the spiral to block diagram, realizes
Show that annual public bicycles user rides data in a limited space.The invention can effectively improve public bicycles system
Understanding of the administrative staff to system operation situation improves the ability of public bicycles data law discovery, is station field signal, vehicle
The decision of scheduling provides foundation.
Description of the drawings
Fig. 1 is the flow chart of visual analysis method of the present invention；
Fig. 2 is that public bicycles user constitutes analysis chart；
Fig. 3 (a) is user type and duration relational graph of riding；
Fig. 3 (b) is the user's amount of riding variation diagram；
Fig. 3 (c) whole year amounts of riding change spiral figure；
Fig. 3 (d) gender calendar figures；
Fig. 4 (a) website spatial distribution maps；
Fig. 4 (b) rides range distribution figure；
Fig. 4 (c) website clustering figures；
Fig. 5 (a) time-space distribution graphs；
The more properties view interaction results of Fig. 5 (b).
Specific implementation mode
As shown in Figure 1, a kind of city public bicycle system borrows the visual analysis method of also pattern, this method specifically to wrap
Include following steps：
Step 1：It collects public bicycles system data and it is pre-processed.
Step 2：Based on user perspective, designing user constitutes analysis chart, not only intuitively shows the composition of user type, simultaneously
Gender and the composition at age are also provided in same figure.
Step 3：Based on time visual angle, user type is devised and duration relational graph of riding, facilitate research different user class
The relationship of type and corresponding duration of riding；The whole year amount of riding variation spiral figure is devised, displaying public bicycles user rides whole year
The changing rule of amount；Design day is ridden variation diagram, the daily changing rule of riding of displaying different user types；It devises and is based on day
The user for going through figure rides ratio chart, the daily ratio ridden of displaying different user types.
Step 4：Based on spatial view, design station space of points distribution map shows the spatial distribution of city public bicycle；If
Meter is ridden range distribution figure, and displaying different user types are ridden the distribution situation of distance；Website clustering figure is designed, to public
The cluster result of bicycle website carries out visual analysis, assists in customer analysis difference and clusters the rule of riding of website, right
The planning of website and the different cluster areas citizens' activities rules of understanding have great importance.
Step 5：More properties views are designed, Main Analysis is under different weather situation, the effect of date property factor, to vehicle
By means of the influence of also quantity.
The step 1 includes：
Step 1.1：Obtain data and Chicago Divvy public affairs in Hangzhou public bicycles on March 23rd, 2014 to June 23
Bicycle system annual data December 31 1 day to 2016 January in 2016 altogether.Two public bicycles system datas all wrap
It includes website essential information and user rides and records information.
Wherein Hangzhou public bicycles system website information table (stationInfo) includes mainly：Site ID
(stationId), site name (stationName), site address (address), latitude (lat), longitude (lng) and website
Stake digit (stallNum).Record sheet (leaseInfo) of riding includes the main information of riding of user, which has：
Bicycle ID (bikeId), it card number (cardNo), Site ID of hiring a car (rentStation), time of chartering (rentTime), hires a car
When can park field (rentBerth), Site ID of returning the car (rentStation), when returning the car the time (returnTime) and returning the car
Can park stake position (returnBerth).Chicago Divvy public bicycles system data set also includes site table and note of riding
Record table.Unlike the public bicycles data of Hangzhou, Chicago Divvy public bicycles system website information tables
(DivvyStations) also include website on-line time other than the main informations such as Site ID, site name, site address
(onlineTime)；Chicago Divvy public bicycles systems ride record sheet (DivvyTrips) also comprising when riding lasting
Between the letters such as (tripduration), user type (usertype), user's gender (gender), year of birth (birthyear)
Breath.
Step 1.2：Data are pre-processed, in the public bicycles data of Hangzhou, record of partly riding website of returning the car is
Null value, it is meant that the bicycle may have been lost；Hangzhou public bicycles ride record in include 48785 numbers in 2012
According to this partial data causes harmful effect to our visual analysis in order to prevent, we directly delete this partial data.
The data that Divvy public bicycles systems provide are ridden record comprising 49 repetitions, we are also deleted；According to Divvy public affairs
24 hours logical card users known to bicycle system data set official explanation do not include gender and age information altogether, therefore delete
24 hours logical card users of 353 band gender informations are ridden record, in order to avoid visual analysis is caused to perplex.According to starting and purpose
The calculation of longitude ＆ latitude of website is ridden apart from (distance), and the translate methods provided using Baidu map API, is realized
Conversion of the GPSS84 coordinate systems to Baidu's coordinate system.
Fig. 2 gives public bicycles user and constitutes analysis chart.The figure is a kind of multi-layer annular figure, in the same figure
The proportionate relationship of multiple levels can be indicated without splitting into multiple pie charts.Different user types uses different colors
It is mapped, the ride number of public bicycles of various user types is mapped to the sector region of the figure, the angle of sector region
Or the bigger amount of the riding proportion for representing the user type of area is also bigger.The figure can not only show different user types
Ratio can also show the concrete numerical value of the number of riding of each user type.Outermost layer maps male and female in year card member
Property age, the colors of age visual elements is mutually unified with the colour system of corresponding gender visual elements, and age bigger color is more
It is deep, meet the cognition custom of people.Public bicycles administrative staff, can the intuitive user's gender and age structure for obtaining understanding system
At.
The step 2 includes：
Step 2.1：Using the interface of Visualization Framework Echarts, according to the statistical for record of riding to public bicycles
Year is blocked the number of riding of member and Tong Ka member by analysis, blocks ride number and the ratio pass of male and female user in member year
System is visualized with circular chart.When user slips over the sector region of corresponding circular chart, it is secondary to pop up riding for the certain customers
Number and proportion.
As shown in Fig. 2, the figure is a kind of multi-layer annular figure, it can indicate that the ratio of multiple levels is closed in the same figure
System is without splitting into multiple pie charts.Different user types is mapped using different colors, and various user types are ridden
The number of row public bicycles is mapped to the sector region of the figure, and the angle or area of sector region are bigger to represent the user type
The amount of riding proportion it is also bigger.The figure can not only show the ratio of different user types, can also show each user
The concrete numerical value of the number of riding of type.Outermost layer maps the age of male and female in year card member, age visual elements
Color is mutually unified with the colour system of corresponding gender visual elements, and age bigger color is deeper, meets the cognition custom of people.It is public
Bicycle management personnel altogether, can the intuitive user's gender and age composition for obtaining understanding system.
The step 3 includes：
Step 3.1：It designs based on the user type for stacking bar chart and duration relational graph of riding.Each bar shaped represents some
It rides the number of riding of user in duration, wherein number of riding includes year card member and Tong Ka member, is divided into not in year card member
The number of riding of male and female user is indicated using blue and red bar shaped, logical card user is indicated using grey.Respectively
The height that user type is stacked is higher, and the amount of riding for representing user in the period is bigger, and vice versa.
Step 3.2：It rides variation diagram in day of the design based on area-graph.In order to bring consistent visual experience, phase to user
It is indicated using identical color with user type.At this point, abscissa indicates that period, ordinate indicate the time in area-graph
The data of riding of the different user types of section, the height of broken line is higher, indicates that date user's amount of riding is bigger.User can lead to
It crosses period selection component and selectes special time period, the data of the period will amplify, assist in user and focus oneself
Interested data, while also contributing to the relationship for helping user to understand the partial data and ambient data.
Step 3.3：Helical structure whole year of the design based on D3.js rides variation diagram, which is saved by way of spiral
Big quantity space, which can not only show the data variation trend of big period, but also can intuitively express data
Periodically.It was indicated using 12 kinds of different colors in 1 year 12 month, contributes to user clearly by the number in different months
According to being distinguished, and it can intuitively observe annual changing rule.User can by click each bar shaped of this figure come
Check the quantity of riding on the date.
Step 3.4：User of the design based on calendar figure rides ratio chart, shows time and the relationship that user is constituted.When with
When family selects some month, which can be precipitated the daily amount of riding of each user type in this month with statistical, and according to than
Example relationship is plotted in the month corresponding calendar.The data of different user types, data volume are mapped using different colors
Bigger shared sector region is bigger.
As shown in Fig. 3 (a), with 6 indexs when user rides：Within 30 minutes, 30 minutes to 1 hour, 1 hour to 2
Hour, 2 hours to 5 hours, 5 hours to 10 hours and 10 hours or more.It, will when mouse is moved on corresponding stacking block diagram
Show the corresponding exact numerical of each user type.Therefrom it will be seen that with duration of riding increase, correspondence ride
The number of riding of period is reduced therewith.Duration ride when year card member's amount of riding within 30 minutes is significantly larger than same ride
Long logical card member；The amount of riding of female user all has been over the amount of riding of logical card user even in year card member.It is riding
In record of riding of the row duration more than 30 minutes, logical card user has overwhelming superiority more than the quantity of riding of year card member.30
In minute to 1 hour duration of riding, the number of riding of male and female user is of substantially equal.
As shown in Fig. 3 (b), more meticulously visual analysis is carried out to public bicycles variations of annual situation.From figure we
As can be seen that most of amount of riding concentrates in April, 2015 between in November, 2016.Winter (January, 2 months and December) is public
Cycling amount is very limited, this is related with the winter weather conditions in Chicago.
In order to which the difference and annual user that preferably show the amount of riding between different months are ridden changing rule, Wo Mentong
Cross the annual global regularity of Fig. 3 (c) displaying amounts of riding.Annual daily data are intuitively presented in by helical design same
In figure.The figure not only facilitates changing rule whole in observation year, is more conducive to observe the data situation in January.
Annular center represents on January 1st, 2016, and the sequence to extend outward from center represents the sequence of time passage.Different colours generation
Table different month, user can clearly distinguish the data in different months, can the sharp situation of change for obtaining perception data.
Gender calendar figure Fig. 3 (d) is designed, the proportionate relationship of riding of different user types in one month is studied.Pass through the figure
Public bicycles user administrator can observe the rule that the different user types amount of riding changes over time.As shown on the right, July
Part gender calendar figure visible weekend leads to the case where card user showed increased is even more than year card user and logical card user on July 4
Abnormal increase.The figure contributes to public bicycles manager analysis various factors to ride different user types user's public bicycles
Capable influence.
The step 4 includes：
Step 4.1：Design station space of points distribution map, the view are that the essential information based on public bicycles system website is looked into
It askes, using polymerization technique, the dense degree of website distribution in certain region is shown according to scaling, website dense degree is reflected
Penetrate the size for radius.The polymerization radius in the more then region of certain region website is bigger, the relatively few then region of website in certain region
Polymerization radius it is smaller.
Step 4.2：Design range distribution figure of riding, data system of the view based on the entire public bicycles system amount of riding
Meter analysis shows the quantity of riding of certain in section different user types of riding using bar chart is stacked.When user passes through
When mouse selects some to stack bar shaped, pop-up shows the amount of riding of the different user types apart from section.
Step 4.3：Design website clustering figure.Using Unsupervised clustering algorithm：AP clustering algorithms (Affinity
Propagation Cluster) clustering is carried out for public bicycles website.
First build similar matrix：
Assuming that there is n website in public bicycles site table, m region is finally required division into.We with S=S1,
S2, S...Sn } indicate public bicycles Website Hosting；{ A1, A2, A3...Am indicate to need the cluster that divides A=.dijTable
Show the space length between any two website i and j, rijIndicate that the public bicycles between website i and j are ridden recorded amounts, i.e.,
Slave site i hires a car and returns the car and leased in website j and the public bicycles given back in website i record summation, wherein i in website j,
j∈S.Thus can all websites of public bicycles be established with space length matrix D is respectively and public bicycles ride and record square
Battle array Rel.
Dis is distance matrix, and setting means is as shown in formula 1.1, wherein dijIndicate website i to the space of website j away from
From.
Rel is record matrix of riding, and set-up mode is as shown in formula 1.2.Wherein rijIndicate the public affairs between website i and website j
Cycling recorded amounts altogether.
The model by public bicycles ride record relational matrix and space length matrix be combined, between structure the website
Similar matrix, in this, as the foundation of cluster.Before the record relationship that will ride is attached to distance matrix as weight, logarithm is needed
According to being normalized.
We are normalized data using such as 1.3 mode of formula, wherein rmaxIndicate the maximum in Rel matrixes
Value, rminIndicate the minimum value in Rel matrixes.By formula it is found that if rent/also relationship between two websites is stronger, wijValue get over
It is small.
By the value w after normalizationijWith website distance value dijThe similar matrix Sn of website can be obtained by blending：
Then the topological structure of entire website is abstracted as graph structure.Then three centrality of figure are calculated, and constitute square
Battle array：
It is finally introducing a variable α and two similar matrix S n and S t is integrated into new similar matrix S i.e. S=α S n+
(1-α)S t。
It is ridden through the above steps according to site location and user and records structure similar matrix, then passed through AP and cluster calculation
Method obtains public bicycles website cluster result.
AP clustering algorithms are as follows：
Step 1:Initialization builds similar matrix s according to the method described above, and initial value is assigned to p.
Step 2:Calculate the value of the representative degree between data point.
Step 3:Calculate the value of the suitable degree of choosing (responsibilities) between data point.
r(i,k)←min{0,r(k,k)+∑j≠i,kmax(0,r(j,k))} (1.7)
a(k,k)←∑j≠kmax(0,r(j,k)) (1.8)
Step 4：To the value of representative degree and suitable degree of choosing (availabilities) to value be updated.
Step 5:If it exceeds the maximum iteration or cluster centre of setting all no longer become in successive ignition
When change, terminates and calculate, the emphasis clustered at this time and the sample point each clustered have determined；Otherwise return to Step2, continue into
Row calculates.
Website clustering figure is finally drawn according to cluster result.Pair of map is mapped to according to the latitude and longitude information of website
Different colors is answered on position and is mapped as the website of different clusters, the mark color of the same cluster website is identical.
Fig. 4 (a) shows the geographical distribution situation of public bicycles website.User type is illustrated using Fig. 4 (b) and is ridden
The relationship of distance, from figure we can see that the distance of riding of most of user in 0.5 km to 5 kms, be less than 0.5
The user of km accounts for small percentage.Wherein most users ride distance in 2 kms between 5 kms.Simultaneously it also seen that having
Only a few user rides distance more than 10 kms.We show that blocking member year averagely rides length as 1.93 kms by statistics；
Logical card user averagely rides length as 2.16 kms.It can reflect the distance slightly above year card of riding that is averaged of logical card user accordingly
Member.By the visual analysis to public bicycles website cluster result (Fig. 4 (c)), it can be found that the spatial distribution to cluster is special
Point and the regularity of distribution of riding in week.It is certain cluster between there are certain general character, then there is larger difference in some, cluster institute between clustering
The property for belonging to region is to lead to the major reason of these similarities and differences.The figure can be that public bicycles system operator carries out website rule
It draws, site zone divides and public bicycles scheduling provides foundation.
The step 5 includes：The more properties views of the design specifically include, and generate influence factor tables of data more than one,
Storage gives the date property and Weather property of every day in analysis time section；Wherein Weather property has seven attribute values：Fine day,
Cloudy, shower, moderate rain, heavy rain, snows at light rain；Wherein date property includes：Working day, weekend, festivals or holidays；Using thermodynamic chart
Displaying different weather and the user under the date ride distribution situation, and user's amount of riding in certain region is bigger, and color is deeper, otherwise also
So；
Design more properties views：It is time shaft assembly below the figure, user selects data set by using time shaft assembly
Date in time range, figure top design weather component, the date selected according to time shaft assembly are inquired weather feelings
Condition, intermediate region are thermodynamic chart, show that the heating power that user rides is distributed according to the date of time shaft selection；
Design multi-attribute analysis figure：The matrix arranged using two rows four is displayed side by side 8 not under same date and weather conditions
Heating power situation；First row is the tourist season, and secondary series indicates that off-season, the first row indicate that working day fine day, secondary series indicate
Working day moderate rain, third row indicate that weekend fine day, the 4th row indicate rainy day at weekend or snowy day.
As shown in Fig. 5 (a), public bicycles in the case of displaying off-season and tourist season, festivals or holidays, different weather
It rides heating power distribution situation.Therefrom find that the website for leading to card user in tourist season temperature of riding is significantly improved and presented and has a rest along close
The characteristics of root lake bank sight spot is distributed.The rule more similar with dull season, but Lincoln Park is presented in the rule of riding for blocking member year
The amount of riding of neighbouring website improves a lot.Block in women year website of the member in depicted area ride it is relatively balanced, in indigo plant
Nearby the public bicycles website amount of riding is significantly improved color subway line Damen Blue and Division subway station.Fig. 5
(b) space distribution rule of weather conditions (containing weather, temperature, the wind-force) and user's amount of riding on June 12nd, 2014 is given
Situation.User can select component, selection to want the period checked and realize by time sliding block to connect by date-time section
The selection on continuous date, top obtains weather condition according to the date that user selects by inquiry, by the user on date selected by user
The space distribution situation of the website amount of riding shows in the map components illustrated again.Public bicycles manager is contributed in depth to divide
Analysis weather and date lease the amount of riding and website of public bicycles user the influence of heating power distribution.
Claims (6)
1. a kind of city public bicycle system borrows the visual analysis method of also pattern, which is characterized in that this method is specifically wrapped
Include following steps：
Step 1：It collects public bicycles system data and it is pre-processed；
Step 2：Based on user perspective, designing user constitutes analysis chart, shows the composition of user type, while in same figure
The composition of gender and age is also provided；
Step 3：Based on time visual angle, designing user type and duration relational graph of riding；It designs the whole year amount of riding and changes spiral figure,
Show the changing rule of public bicycles user's whole year amount of riding；Design day is ridden variation diagram, and displaying different user types are daily
Changing rule of riding；It devises the user based on calendar figure to ride ratio chart, displaying different user types are daily to ride
Ratio；
Step 4：Based on spatial view, design station space of points distribution map shows the spatial distribution of city public bicycle；Design is ridden
Row distance distribution map, displaying different user types are ridden the distribution situation of distance；Design website clustering figure, to it is public voluntarily
The cluster result of station point carries out visual analysis；
Step 5：More properties views are designed, analysis borrows also quantity under different weather situation, the effect of date property factor, to vehicle
Influence.
2. a kind of city public bicycle system according to claim 1 borrows the visual analysis method of also pattern, special
Sign is：The collection public bicycles system data includes that website essential information and user ride and record information；Wherein stand
Point essential information includes Site ID, site name, site address, latitude, longitude, website on-line time and website stake digit；Institute
The user stated record information of riding includes bicycle ID, card number, Site ID of hiring a car, time of chartering, the field that can park when hiring a car, also
Vehicle Site ID, time of returning the car, the when of returning the car can park stake position, duration of riding, user type, user's gender and year of birth,
Wherein user type includes logical card user and year card user, and the record of riding of middle aged card user includes gender and age, is led to
It does not include gender and age in the record of riding of card user；The pretreatment is：It is riding for null value to delete website of partly returning the car
Row record, deleting duplicated data and the data other than effective time, it is sky to delete ride gender or age in record of year card user
The data of value are ridden distance according to the calculation of longitude ＆ latitude of starting and purpose website, are provided using Baidu map API
Translate methods realize conversion of the GPSS84 coordinate systems to Baidu's coordinate system.
3. a kind of city public bicycle system according to claim 1 borrows the visual analysis method of also pattern, special
Sign is：Step 2 specifically, using Visualization Framework Echarts interface, according to the statistics for record of riding to public bicycles
Male and female user in ride number and the proportionate relationship of blocking member and Tong Ka member year, year card member is ridden in analysis
Number and proportionate relationship, different age group ride number and proportionate relationship is visualized with circular chart；Corresponding circular chart
Sector region includes ride number and the proportion of the certain customers.
4. a kind of city public bicycle system according to claim 1 borrows the visual analysis method of also pattern, special
Sign is：Step 3 is specifically, user type and ride duration relational graph of the design based on stacking bar chart；Each bar shaped represents
The number of riding of user in some duration of riding blocks in member again in year wherein number of riding includes blocking member and Tong Ka member year
Two different color bar shapeds are respectively adopted to indicate the number of riding of male and female user, leads to card user and uses with above-mentioned two
The third color other than color is planted to indicate；The height that each user type is stacked is higher, represents user in the period
The amount of riding is bigger, and vice versa；
The wherein annual amount of riding variation spiral figure was variation diagram of riding helical structure whole year of the design based on D3.js, by 1 year
12 month was unfolded using 12 kinds of different colors come spiral, and the quantity of riding on the date is checked in each bar shaped；
It rides variation diagram in day of the design based on area-graph；Wherein different user types use and user type and duration relationship of riding
Identical Color scheme is schemed to indicate；Abscissa indicates that period, ordinate indicate the different user of the period in area-graph
The data of riding of type, the height of broken line is higher, indicates that date user's amount of riding is bigger；User selects component by the period
Selected special time period, the data of the period will amplify；
User based on calendar figure rides ratio chart, shows time and the relationship that user is constituted；When user selects some month,
The amount of riding that each user type is daily in this month is precipitated in the view statistical, and is plotted to month correspondence under a proportional relationship
Calendar in；The data of different user types are mapped using different colors, data volume is bigger, and shared sector region is bigger.
5. a kind of city public bicycle system according to claim 1 borrows the visual analysis method of also pattern, special
Sign is：
The website spatial distribution map, which is the essential information inquiry based on public bicycles system website, using poly-
Conjunction technology shows the dense degree of website distribution in certain region according to scaling, website dense degree is mapped as radius
Size；The polymerization radius in the more then region of certain region website is bigger, the polymerization half in the relatively few then region of website in certain region
Diameter is smaller；
The range distribution figure of riding, data statistic analysis of the view based on the entire public bicycles system amount of riding, is adopted
The quantity of riding of certain in section different user types of riding is shown with stacking bar chart；
The design website clustering figure；Using Unsupervised clustering algorithm：AP clustering algorithms are for public bicycles website
Carry out clustering；
First build similar matrix：
Assuming that there is n website in public bicycles site table, m region is finally required division into；S={ S1, S2, S...Sn } comes
Indicate public bicycles Website Hosting；A={ A1, A2, A3...Am } indicates to need the cluster that divides；dijIndicate any two
Space length between website i and j, rijIndicate that the public bicycles between website i and j are ridden recorded amounts, i.e. slave site i hires a car
And it returns the car and is leased in website j and the public bicycles given back in website i record summation, wherein i, j ∈ S in website j；To public
All websites of bicycle establish space length matrix D is respectively and public bicycles ride and record matrix Rel；
Dis is distance matrix, and setting means is as shown in formula 1.1, wherein dijIndicate the space length of website i to website j；
Rel is record matrix of riding, and set-up mode is as shown in formula 1.2；Wherein rijIndicate website i and website j between it is public from
Driving is ridden recorded amounts；
The model by public bicycles ride record relational matrix and space length matrix be combined, it is similar between structure the website
Matrix, in this, as the foundation of cluster；Before the record relationship that will ride is attached to distance matrix as weight, need to data into
Row normalized；
Data are normalized using such as 1.3 mode of formula, wherein rmaxIndicate the maximum value in Rel matrixes, rminTable
Show the minimum value in Rel matrixes；By formula it is found that if rent/also relationship between two websites is stronger, wijValue it is smaller；
By the value w after normalizationijWith website distance value dijThe similar matrix Sn of website can be obtained by blending：
Then the topological structure of entire website is abstracted as graph structure；Three centrality of figure are calculated, and constitute matrix：
A variable α is finally introducing by two similar matrix SnAnd StIt is integrated into new similar matrix S i.e. S=α Sn+(1-α)St；
It is ridden through the above steps according to site location and user and records structure similar matrix, then obtained by AP clustering algorithms
Public bicycles website cluster result；
Website clustering figure is finally drawn according to cluster result；The correspondence position of map is mapped to according to the latitude and longitude information of website
Different colors is set and is mapped as the website of different clusters, the mark color of the same cluster website is identical.
6. a kind of city public bicycle system according to claim 1 borrows the visual analysis method of also pattern, special
Sign is：
The more properties views of the design specifically include, and generate influence factor tables of data more than one, store and give analysis time section
The date property and Weather property of interior every day；Wherein Weather property has seven attribute values：Fine day, cloudy, shower, light rain, in
Rain, snows at heavy rain；Wherein date property includes：Working day, weekend, festivals or holidays；Different weather and date are shown using thermodynamic chart
Under user ride distribution situation, user's amount of riding in certain region is bigger, and color is deeper, and vice versa；
Design more properties views：It is time shaft assembly below the figure, user selects data set time by using time shaft assembly
Date in range, figure top design weather component, the date selected according to time shaft assembly are inquired weather condition, in
Between region be thermodynamic chart, show that heating power that user rides is distributed according to the date of time shaft selection；
Design multi-attribute analysis figure：The matrix arranged using two rows four is displayed side by side 8 not heating power under same date and weather conditions
Situation；First row is the tourist season, and secondary series indicates that off-season, the first row indicate that working day fine day, secondary series indicate work
Day moderate rain, third row indicate that weekend fine day, the 4th row indicate rainy day at weekend or snowy day.
Priority Applications (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810101548.3A CN108470033A (en) | 2018-02-01 | 2018-02-01 | A kind of city public bicycle system borrows the visual analysis method of also pattern |
Applications Claiming Priority (1)
| Application Number | Priority Date | Filing Date | Title |
|---|---|---|---|
| CN201810101548.3A CN108470033A (en) | 2018-02-01 | 2018-02-01 | A kind of city public bicycle system borrows the visual analysis method of also pattern |
Publications (1)
| Publication Number | Publication Date |
|---|---|
| CN108470033A true CN108470033A (en) | 2018-08-31 |
Family
ID=63266029
Family Applications (1)
| Application Number | Title | Priority Date | Filing Date |
|---|---|---|---|
| CN201810101548.3A Pending CN108470033A (en) | 2018-02-01 | 2018-02-01 | A kind of city public bicycle system borrows the visual analysis method of also pattern |
Country Status (1)
| Country | Link |
|---|---|
| CN (1) | CN108470033A (en) |
Cited By (7)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN109741534A (en) * | 2019-01-18 | 2019-05-10 | 绵阳市智慧鸟软件技术有限公司 | A kind of vertical lift picking mechanism of vending machine |
| CN109903125A (en) * | 2019-01-28 | 2019-06-18 | 东南大学 | Shared bicycle based on OD data borrow also with park spatial and temporal distributions method for visualizing |
| CN111397617A (en) * | 2019-01-03 | 2020-07-10 | 阿里巴巴集团控股有限公司 | Method and device for planning path for specific vehicle |
| CN112184088A (en) * | 2020-11-12 | 2021-01-05 | 同济大学 | Urban ride quality evaluation method and system based on multi-source data |
| CN112699284A (en) * | 2021-01-11 | 2021-04-23 | 四川大学 | Bus stop optimization visualization method based on multi-source data |
| CN113689124A (en) * | 2021-08-26 | 2021-11-23 | 悉地（苏州）勘察设计顾问有限公司 | Internet rental bicycle demand characteristic evaluation method and storage medium |
| CN117573953A (en) * | 2024-01-16 | 2024-02-20 | 成都云祺科技有限公司 | Page big data visual rendering method, system and storage medium |
Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105205623A (en) * | 2015-10-29 | 2015-12-30 | 杭州电子科技大学 | Public bicycle station dispatch area division method based on interval weak coupling degree |
| CN105321226A (en) * | 2014-07-31 | 2016-02-10 | 福特全球技术公司 | Method and system for vehicle data gathering and analysis |
| CN106296350A (en) * | 2016-08-04 | 2017-01-04 | 杭州电子科技大学 | A kind of visual analyzing city public bicycle system borrows the method for also pattern |
| CN107423852A (en) * | 2017-07-24 | 2017-12-01 | 华北电力大学（保定） | A kind of light storage combined plant optimizing management method of meter and typical scene |
-
2018
- 2018-02-01 CN CN201810101548.3A patent/CN108470033A/en active Pending
Patent Citations (4)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN105321226A (en) * | 2014-07-31 | 2016-02-10 | 福特全球技术公司 | Method and system for vehicle data gathering and analysis |
| CN105205623A (en) * | 2015-10-29 | 2015-12-30 | 杭州电子科技大学 | Public bicycle station dispatch area division method based on interval weak coupling degree |
| CN106296350A (en) * | 2016-08-04 | 2017-01-04 | 杭州电子科技大学 | A kind of visual analyzing city public bicycle system borrows the method for also pattern |
| CN107423852A (en) * | 2017-07-24 | 2017-12-01 | 华北电力大学（保定） | A kind of light storage combined plant optimizing management method of meter and typical scene |
Non-Patent Citations (2)
| Title |
|---|
| 杨欢欢 等: "基于螺旋图的时间序列数据可视化", 《计算机应用》 * |
| 陈红倩 等: "一种针对农残检测数据的时序分组可视化方法", 《系统仿真学报》 * |
Cited By (12)
| Publication number | Priority date | Publication date | Assignee | Title |
|---|---|---|---|---|
| CN111397617A (en) * | 2019-01-03 | 2020-07-10 | 阿里巴巴集团控股有限公司 | Method and device for planning path for specific vehicle |
| CN111397617B (en) * | 2019-01-03 | 2023-09-22 | 阿里巴巴集团控股有限公司 | Method and device for planning path for specific vehicle |
| CN109741534A (en) * | 2019-01-18 | 2019-05-10 | 绵阳市智慧鸟软件技术有限公司 | A kind of vertical lift picking mechanism of vending machine |
| CN109741534B (en) * | 2019-01-18 | 2024-06-07 | 广东柏菲智能科技有限公司 | Vertical lifting goods taking mechanism of vending machine |
| CN109903125A (en) * | 2019-01-28 | 2019-06-18 | 东南大学 | Shared bicycle based on OD data borrow also with park spatial and temporal distributions method for visualizing |
| CN112184088A (en) * | 2020-11-12 | 2021-01-05 | 同济大学 | Urban ride quality evaluation method and system based on multi-source data |
| CN112184088B (en) * | 2020-11-12 | 2022-05-17 | 同济大学 | Urban ride quality evaluation method and system based on multi-source data |
| CN112699284A (en) * | 2021-01-11 | 2021-04-23 | 四川大学 | Bus stop optimization visualization method based on multi-source data |
| CN112699284B (en) * | 2021-01-11 | 2022-08-30 | 四川大学 | Bus stop optimization visualization method based on multi-source data |
| CN113689124A (en) * | 2021-08-26 | 2021-11-23 | 悉地（苏州）勘察设计顾问有限公司 | Internet rental bicycle demand characteristic evaluation method and storage medium |
| CN117573953A (en) * | 2024-01-16 | 2024-02-20 | 成都云祺科技有限公司 | Page big data visual rendering method, system and storage medium |
| CN117573953B (en) * | 2024-01-16 | 2024-03-22 | 成都云祺科技有限公司 | Page big data visual rendering method, system and storage medium |
Similar Documents
| Publication | Publication Date | Title |
|---|---|---|
| CN108470033A (en) | A kind of city public bicycle system borrows the visual analysis method of also pattern | |
| Brown et al. | An analysis of the relationships between multiple values and physical landscapes at a regional scale using public participation GIS and landscape character classification | |
| Huang et al. | A global comparative analysis of urban form: Applying spatial metrics and remote sensing | |
| Brown et al. | The extrapolation of social landscape values to a national level in New Zealand using landscape character classification | |
| CN108280550B (en) | Visual analysis method for comparing community division of public bicycle stations | |
| Skinner | Cities and the hierarchy of local systems | |
| CN109074622A (en) | Determine the system and method for transportation service route | |
| Burns | Tourism in Russia: background and structure | |
| Green | Coastal towns in transition: Local perceptions of landscape change | |
| Cohen et al. | Form and function in the geography of retailing | |
| Kraas et al. | Socio-economic atlas of Myanmar | |
| Ke et al. | Contribution of non-park green space to the equity of urban green space accessibility | |
| Wibowo | Interval-valued intuitionistic fuzzy multicriteria group decision making approach for hotel selection | |
| Li et al. | Multiscale spatiotemporal dynamics analysis of urban green space: Implications for green space planning in the rapid urbanizing Hefei City, China | |
| Luo et al. | Potential recreation service efficiency of urban remnant mountain wilderness: A case study of Yunyan District of Guiyang city, China | |
| Okafor | The spatial dimensions of accessibility to general hospitals in rural Nigeria | |
| Williamson et al. | Visual resource management of Victoria's forests: a new concept for Australia | |
| Betz et al. | Outdoor recreation resources | |
| CN114331232A (en) | Street space quality monitoring, evaluating and early warning method | |
| Cai et al. | The coupling coordination between tourism urbanization and ecosystem services value and its obstacle factors in ecologically fragile areas: a case study of the Wuling Mountain area of Hunan Province, China | |
| Devi et al. | Measuring urban form units: Alternative for characterizing urban growth pattern in Yogyakarta Urbanized Areas | |
| Lloyd et al. | Measuring local segregation in Northern Ireland | |
| Pranke et al. | Medieval trade in Central Europe, Scandinavia, and the Balkans (10th-12th centuries): a comparative study | |
| Tanser | The application of geographical information systems to infectious diseases and health systems in Africa. | |
| Aytekin et al. | Determination of outdoor recreation potential: case of the city of Bartin and its environs, Turkey |
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 | ||
| RJ01 | Rejection of invention patent application after publication | ||
| RJ01 | Rejection of invention patent application after publication |
Application publication date: 20180831 |