mobmap hands-on @ FOSS4G SEOUL 2015

FOSS4G 2015 14 Sep 2015

Hiroaki Sengoku, Ph.D Satoshi Ueyama

Ritsu Sakuramachi

Mobmap: Introduction to People Flow Analysis

Purpose and Summary

This course covers how to visualise and analysis time-series data such as trajectory data using Mobmap for beginners. In this course we will use simulated people flow data developed by The University of Tokyo, Center for Spatial Information Science(CSIS).

About microbase inc.microbase Inc. is the company which creates micro demographic data in Japan. This company has created simulated urban data such as people flow or people life style using open data. The member of microbase Inc. aim to create micro demographic data all over the world and simulation platform such as Sim City using these data.

Real-estate values Building Age

Future PopulationMicro CensusPersonal LifeStyle

Team mobmap

~!!

ADENDA

SESSION 1 INTRODUCTION Mobmap Hands-on data

SESSION 2 ANALYSIS Basic operation Attribute query Spatial query Other functions

SESSION 3 PRACTICE Practice Summary

SESSION 1 INTRODUCTION

Where do visitors come from?

Software used this Hands-on

Mobmap (Google Chrome)

QGISFree

Google Earth (Free)

20131123 16:18:21 35.979504,139.827072

Visualising Moving Data, InteractivelyUsing Mobmap for Google Chrome

Satoshi Ueyama EDITORIA, The University of Tokyo

http://shiba.iis.u-tokyo.ac.jp/member/ueyama/mm/

http://shiba.iis.u-tokyo.ac.jp/member/ueyama/mm/

Whats mobmap?

Tool for visualising and analyzing time-series data (ex.GPS log) Show animation for time-series change on Google Maps Windows, Mac, Linux supports

1500,1,1,1998/10/01 06:00:00,139.9249985549,35.7318406842,2,7,4110309,14,97,33,,973700,1,1,1998/10/01 06:00:00,139.9123053021,35.753511987,1,10,4112107,10,97,33,,977300,1,1,1998/10/01 06:00:00,139.9132597066,35.7134959947,1,7,4114009,8 ,97,40,,97

5500,1,1,1998/10/01 06:00:00,139.9374260851,35.7387718937,2,12,4113004,14,97,32,,979500,1,1,1998/10/01 06:00:00,139.9268670539,35.6868715236,1,2,4115011,12,97,26,,979700,1,1,1998/10/01 06:00:00,139.9238668934,35.6892555155,2,6,4115016,14,97,32,,9711400,1,1,1998/10/01 06:00:00,139.9293917865,35.6808909812,1,6,4115107,9 ,97,36,,9711800,1,1,1998/10/01 06:00:00,139.9077829215,35.6792209637,2,6,4115202,14,97,21,,9710100,1,1,1998/10/01 06:00:00,139.9298447577,35.684551261,1,1,4115014,12,97,26,,97

Whats mobmap?

Input Output

time-series data(CSV)Route data(KML)

Movie(MP4)Mesh data(CSV)

Polygon data(KML)

Install

Install Google Chrome (if you don't install it ) Search "mobmap2" in Chrome Web store

Application start

Open a new tab of Google Chrome Open Google Chrome launcher

Open Mobmap

Target area

Whole Tokyo (except for an isolated island

Simulated People Flow data

Kansai, Nakagyo version also have been released.

Data used this hands-on

Simulated People Flow dataas time-series data

2013-07-01.csv 2013-07-07.csv2013-10-07.csv 2013-10-13.csv2013-12-16.csv 2013-12-22.csv



Metropolitan

Chukyo

Kansai

Data used this hands-onCommercial accumulation statisticsPolygon data

ca_2011_13.kmlca_2011_23.kmlca_2011_27.kml

TokyoNagoyaOsaka

Stay population data(pstay)

pstay_sample.csv

Practice dataSimulated People Flow data


Simulated People Flow data is made from geo-tagged Tweet data(presented by Nightlei Co., Ltd.)

I'm at Ramen Jiro Meguro shop (Meguro-ku)

139.70714271068635.6341373645078

ex)


This data is created as following estimation and interpolation methods from geo-tagged Tweet data.

Home estimate

Stay time estimation

Path interpolation

Home place is defined as a city and district which users have frequently checked in on morning and a holiday. Finally, the place is determined at random in the city.

Virtual stay time is set in advance per category of the check-in (movie, amusement, etc)

Paths are interpolated based on the places between check-in places using road data (cooperation: Hiroshi kanasugi, People Flow Team at Tokyo University CSIS)

Raw geo tagged tweet data on a map (without the interpolations)

Step1

Home place and stay time are given to geo tagged tweet data according to the check-in on map (night-time).

Step2

Virtual stay time per about 250 check-in place category

Paths are interpolated (only in the road) for creating Simulated People Flow data per 5 min using INFORMATION PLATFORM FOR PEOPLE FLOW ANALYSIS by the university of Tokyo CSIS.

Step3

"STUDY OF INFORMATION PLATFORM FOR PEOPLE FLOW ANALYSIS IN URBAN AREA", the 36th Japan Society of Civil Engineering information use technology symposium, pp.111-114, 2011 about Yoshihide Sekimoto, your Satoshi Usui Hiroshi kanasugi, Yusuke Masuda,


id sex date lat lon category1 category2 mode category

105 male2013-07-01 22:10:39 35.71899231 139.31707368 MOVE

105 male2013-07-01 22:15:39 35.71513008 139.31903984 MOVE

105 male2013-07-01 22:20:39 35.71300252 139.31492206home arrival MOVE 8

105 male2013-07-01 22:25:39 35.71483377 139.31029481

arts_entertainment Art Gallery MOVE 4

105 male2013-07-01 22:30:39 35.71591093 139.30722089home arrival STAY 8

1071male2013-07-01 00:00:00 35.72355807 139.73582609home departure STAY 8

Following four attributes are necessary for using Mobmap. "id" (user ID), "date" (time information), "lat", "lon"

Practice dataSimulated People Flow Path interpolation of the practice data is given only data of "MOVE" (during movement), and it is interpolated for every 5 minutes. The railroad network is not reflected by course interpolation.

Other user

Time yyyy-mm-dd HH:MM:SS

Category of the stay spot Detailed category of the stay spot

The information such as twitter id deleted it from the viewpoint of privacy protection

id sex date lat lon category1 category2 mode category

105 male2013-07-01 22:10:39 35.71899231 139.31707368 MOVE

105 male2013-07-01 22:15:39 35.71513008 139.31903984 MOVE

105 male2013-07-01 22:20:39 35.71300252 139.31492206home arrival MOVE 8

105 male2013-07-01 22:25:39 35.71483377 139.31029481

arts_entertainment Art Gallery MOVE 4

105 male2013-07-01 22:30:39 35.71591093 139.30722089home arrival STAY 8

1071male2013-07-01 00:00:00 35.72355807 139.73582609home departure STAY 8

For people to want to play with more Simulated People Flow data

http://www.cs.uic.edu/~wolfson/html/p2p.html

http://research.microsoft.com/apps/pubs/?id=152883

University of Illinois Chicago school (around Illinois)

Microsoft Research(around Beijing)

Practice Commercial accumulation dataPolygon data

Estimated commercial area such as downtown from yellow page which Zenrin Co., Ltd. offers by Yuki Akiyama, a researcher at the university of Tokyo CSIS.

Practice Commercial accumulation dataPolygon data

Researchers can use it under collaborative research with the university of Tokyo from (JORAS)

Unit: Prefecture unit (all over Japan)

Time: 2010 2011

Practice dataTransient population data

Transient population data of the stores around Yoyogi-Uehara Station (Japan) is created using crowdsourcing applications by the PStay project , a crowd souring project at micro geo data workshop.

The PStay project collects the transient population of a place and quantity of traffic, the parking number by crowdsourcing.

http://geodata.csis.u-tokyo.ac.jp/mgd/?page_id=926

SESSION2 WORKSHOP

MobmapBasic operation

Attribute query

Spatial query

Read data Several visualisation

Multi conditions

Line gate Polygon gata

Animation GoogleEarth

Export

Other visualisation Mesh visualisation 3D view

Basic Operation

Read data 1

Choose "Moving Objects" among a button forming a line in the welcome page and open the CSV file

Read data 2

Before loading data, mobmap shows a preview of data When the data include lonlat located in Japan, the lonlat columns are

automatically selected.

You click and change a column as necessary column.

You can change a column when clicking

Read data 3

Click "Start loading" of the lower part. Without any errors, Mobmap starts reading all data.

Start reading in

Practice

Read Simulated People Flow data, Modify!

Practice

Well read on map?

PracticeDate changes

There are "Play", "Stop", "forwarding" button like a movies player. Each object begins to move when the Play button is clicked.

Layer list

Add the layer that was formed by Read data to a list of layers of the left pane

The movable thing can replace order

Additional layer

Layer setting

You can select detailed setting including the indication method of the layer in a list of layers.

Change order of layers Display layer Delete layer

Read Polygon KML

Add the layer from the drop-down menu The polygon supports only KML and WGS84

sample data commercialDistricts.kml

Read Polygon KML

Display the polygon data

Read data 4 (application)

When you want to load an another attribute excepting basic attributes, input "a field name : data type" in the additional line.

Enter category:int

Read data 5 (application)

To change Marker option, choose "By attribute" and change a field name in Vary by attribute.

Labeling function

Labeling attribute data

Turn on a check and appoint Attribute

Tail function

Add a tail depending on movement speed

Attribute

Path Visualisation 1

Choose Target from drop-down menu of Settings

Click Settings

Click Settings


A movement course is visualised by each user


Change the color

Change Dark color

Check


visualise people flow at a specified time

2.Drag to choose time span

1.Click

Tips: when you choose time span, press-and-hold Shift and drag, and you can get regular time.

Time span selection


People flow from 10:00 a.m. to 12:00 a.m

Attribute tableClick Data tab

Display an attribute of each person

Attribute table

Display an attribute of each person

Change a table to display

Practice

Read Transient population data (pstay_sample.csv) and Display it separating by color. A line called "est_pop" shows population per minute.

Reading time series data

Were you able to display it when you changed time so that the color of the marker changed?

Symbol size emphasis

A marker can change its color and size depending on the attribute per minute

Change Markers presets to "Large Scaling Marker"

Attribute query

Attribute query

enables mobmap to choose data under a specific condition

Attribute query

Enter field name = level

category=4 Enter

1retail store (various) 2traffic 3restaurant 4entertainment, leisure 5retail store (food) 6Education 7Other 8Home

Category

Attribute query

Choose an object in accord with a condition

Deselecting

Deselect from an button in the screen of the top left

High property search

||(OR)&&(AND)

In the case of plural conditions

category=4 || category = 8

OR sentenseor

AND sentenseand

category 1

Spatial Query

About spatial query

Deselecting

select of the polygon

select of the rectangle

select of the line gate

Choose a movement object from the select button of the upper part menu

Basic selection

Choose data drawing a rectangle on map

Drug

Basic selection

Only selected area are colored.

Multiple selection

and / or can be used after selecting one objects.

or andnew

Selection mode Already selected

select other areas

andor

Multiple selection

can choose multi area or only overlapped area

OR AND

Invert SelectionClick

Selected area can be inverted.

Gate function

Choose a person, a thing via a certain spot Line gate (appoint it in a segment of a line) Polygon gate (appoint it in a domain)

Line gate application example

Extract a user of Soubu Line

Mount a gate in Shinkoiwa

Line gate application 1

Apply to the expressway along Haneda Airport

click a line gate button

it can pull a line when drag it over a map.


The details are coordinated by a line choice option

After pull a line, a menu is displayed by the line upper part

OK button Direction choice (up, down, both)

Bookmark of lineCancel


35 objects was chosen


Path visualisation also changes under the condition.

Practice

Analyse passing traffic with a line gate function at any place.

Polygon gate Function

Only a point inside polygons

include the trace inside polygons

Polygon gate application 1

Choose the polygon data of the commerce accumulation data that weve read it before.


The attribute of the chosen polygon is shown

choose a polygon layer in a combo box


A detail menu of polygon can be shown when a line of the polygon ID is clicked

Indication of the choice polygon (in a map)

Single choiceDeselecting

Polygon gate function

After having developed the line of the table, click a button


Choose only the movement object which passes a polygon by choosing a button "point + edge" or points only"


Similarly, only the selected path can be visible

Spatial-Attribute query

Possible to combine spatial query and attribute query together

Input an form

Buffer based spatial query 11.Click

create a buffer polygon from an annotation point

2.Put a marker on map

Buffer based spatial query 2

create a buffer polygon from an annotation point

3.Input radiuse

Buffer based spatial query 3

Similarly, the moving objects can be caught with a buffer polygon

A buffer layer is added

Practice

Analyse the visitor of any commerce accumulation place.

Export

ExportThese data can export this data from here.

Without selection, all data be exported

KMLexport

An export button displays it on an attribute table if an object was chosen with a line gate or a polygon gate

Export button

KMLexport

Export the movement course of a chosen movement object in KML form as a polyline

KMLexport

It is saved by default by a name "mobmap-exported.kml"

Coopration with other software

Be careful the attribute because only the first record is reflected If it is sex not to change in time series, there is no problem

Read QGIS

Cooperation with QGIS

Display in QGIS. Read KML form.


It is compatible with other layers


Because Mobmap is specialized in the visualising and analyzing moving trace data, the operation of the general GIS is carried out on QGIS

Example Coordinate transformation (cases of the rectangular coordinates system plane a file)

File conversion to KML form, CSV form Space analysis, operation such as the buffering

Cooperation with GoogleEarth

Of course the data is available in Google Earth

Animation Export functionAdjust screen

adjust a screen for the animation export.

Please put time bars together at the time when you want to start an animation.

If animation export preparations are possible, and then click this button.

Animation export functionChoose an output screen

choose Window of "Mobmap for Chrome"

Animation export functioninput output information

Detailed setting of the animation to output

Output size

set it from here to raise flame

By the default setting, output a share for ten minutes in animation reproduction one second. eg: In the case of 15sec, it is 150min

Animation export functioninput output information

Title

Display a clock

Detailed setting for Movie Export

Animation export functionEntry the output data

Output range can be coordinated by adjustment of the screen size

Animation export functionStart output

After setting up, start exporting movies

with "Record" button

Animation export functionPreserve animation

Save it with "Save" button

Check the animation

the sample movie can be seen on YouTube URLhttp://youtu.be/iAqJR-cy73M

http://youtu.be/iAqJR-cy73M

Other visualisation

Visualising Mesh data

2.Choose Mesh CSV and open the CSV fileNationalCensus__3JTokyo-2010.csv or NationalCensus_4JTokyo-2010.csv.

1.Click

Ex. National Population Census in Tokyo


Click Cell appearance

Mobmap data change Mesh data


1.Click Choose preset

2.Choose Heat (transparent)

Mesh data can separate by color


Much population is red, and Little is blue

Visualising night-time population

2.Click 3.Choose CSV File

1.Click Statiscal operation


Check boxes

Choose AVG

AVG function calculates average of attribute values


The ratio of each local night-time population are shown. The data in this hands-on doesn have the magnification factor and

completeness of parameter so the ratio is 0% largely.


With Person Trip data, Mesh data becomes like this


COUNT function just counts points inside the cell

Chose COUNT


SUM function sums attribute values up

Chose SUM

3D View

Click

Watch 3D map of this data

To use this function, you have to select data in advance.

3D View

A new window will be shown

3D View

You can check time series variation and an abnormal movement easily.

SESSION3 PRACTICE

Practice

Using Simulated People Flow data, decide the target area and find the characteristic trend of the place and consider the reason. Finally, have an effective presentation using mobmap movie function.

PresentationPlease upload a movie which you tried in practice as an animation in YouTube. After creating the movie, tell us the movie URL.

SummaryUsing Mobmap, We learned the method to visualise and analyze GIS data with the time-series data.

This exercise provide for simulated people flow data as sample data. Also, you can handle your own data as well.

Request of the questionnaire

http://goo.gl/bwgCcl

Please answer following questionnaire to improve Mobmap and hands on in future.

Thanks

For creating data and this exercise, Hiroshi Kanasugi helped us to interpolate and create the Simulated People Flow data. The Simulated data is made from the geo tagged tweet data by Ishikawa, Nightlei Co., Ltd.. We appreciate them.

Reference

Reading mesh CSV

Only CSV is the correspondence in the current version. It is not for analysis but for drawing.

Sample data Census-MeshTest2005_3.csv

@static-mesh

@use-mesh-code 3

36533748 0

49395673 0

51394139 0

53393642 0

53393653 0

process a format as follows to display it in mobmap.The first line describe it in the first row with "@static-mesh"The second line describe "@ use-mesh-code" in the first row describe a scale of the mesh in the second row ex) In the case of the third mesh -> 3After the third line

Value (population)Mesh code

Reading in typhoon data

It reads in the behavior of the typhoon from the website

Source: degital typhoon data


Enter the URL of the typhoon page of the digital typhoonhttp://agora.ex.nii.ac.jp/digital-typhoon/summary/wnp/s/201115.html.ja


Display the movement trace of the typhoon with an animation

mobmap hands-on @ FOSS4G SEOUL 2015

Data & Analytics

Transcript of mobmap hands-on @ FOSS4G SEOUL 2015