Min Seok Won

University of Seoul

I want to calculate network centrality index. But currently my analysis unit is link, so I cannot match typical centrality index which is calculated by node with feature ID.

So while I was studying about centrality measures, I found that there is a name "edge centrality" which is for centrality of links. And I tried to find libraries that I can use in python or R, but nothing except betweenness...nevertheless I want to analyze further measures like eigenvector centrality, degree centrality and so on. Since I am not good at coding, I want to use library which is provided in the analysis platform if I can. Is there any edge centrality library except edge betweenness?

Hello I am about to conduct a study related to extracting quarantine point when contagious disease occurred. In my study, I will conduct quantitative analysis, not quantitative analysis.

Thus, I have to make a model, and guess either of following two would be the objective function.

1) classifying the analysis unit(point, road link, small area whatever is ok, but not a region because analysis unit of the model should be small) into one that need to be quarantined and don't have to be quarantine or,

2) extracting which analysis unit is relatively fragile and strongly exposed to risk of infection

I wonder if there's any study about above. And what I wanna mainly discover with those are 'variable' and 'criteria value(threshold)' input for the modeling process.

I've searched very much times, but couldn't find meaningful one and it is very difficult because this is beyond my domain knowledge.

So I would be very please if you help me. Thank you.

Hello, I'm an undergraduate conducting a study about the transport.

While doing this, I recently faced on some questions that bothering me so far.

Of course I've studied about such notions like digital twin, metaverse, vr, ar, and something like that and I've got that the digital twin let us simulate and predict future and apply the best approach to solve the problem.

But still I don't know to how much degree it can influence. I guess this is because I haven't seen the user interface of digital twin till now...

For example, are we able to influence reality right through the digital twin? In other words, under the digital twin platform can we operate and change the real world enviornment by modifying the preset value calculated by the platform? I mean I wonder if we can influence by just clicking the button and modifying the digital twin, not the real world module of infrastructure at all.

And plus, if so(can control the reality by just modifying the digital twin), can it be fully automized? Thus, can a digital twin world conduct following procedures by itself

'find problems of the system - shut down the system or initiate the backup scenario - meanwhile iterate the simulation about the problem - find the best scenario - transmit it to real world module - reinitiate so that the scenario can be operated with the optimized scenario'

Or does 'with digital twin we can optimize and predict future' mean that we can find the best and efficient solution with digital twin, but have to modify real world component if we wanna apply it?

I know that with multimomial logit model I can calculate a probability of each altanatives, and to do this, utility function that consists some variables should be identified. But the problem is now.

As I know, the process that finding the proper value of parameter of variables in utility function is called 'calibration'. Then how can I calibrate these? There're some papers about which of this is mentioned on but for me it is somewhat difficult to understand.

So I would be very appreciate if you tell me clearly, how to do calibration with my data step by step.

I want to make a map that consists of nodes and links in GIS(if in QGIS it'll be better but ArcGIS is also ok).

The information that I only have is this csv file, that contains the link object ID, coordinates of two points of each link, and the uclidean length. You can ignore 'connectivity' tab.

In this situation, can I make the map? If so, how can I? Please please help me :(

I ran the graph analysis with my axial map in depthmapX, which outcomes some properties like controllability, integration and so on. And I reviewed the scatter plot and it's regression equation between two variables to figure out whether I conducted the analysis correctly.

Of course you might've known, in 'scatter plot' menu, we can check the scatter plot, and the R^2 and regression equation of it.

But the matter is, among these scatter plot and R^2 seem reasonable, but not the regression equation. I mean, regard to equation, the outcome shows the result that is impossible to be outcome...It is easy to know that because in all equation of scatter plots that visualize 'any' two variables, intercept and parameter B are 'exactly' same. For example, y = ax + a / y=bx + b ...not the shape of y = ax+b. It could be easier to understand what I'm talking about if you see the right upper side of the picture that I uploaded.

So could you tell me why this happen and how to solve this problem, or

tell me how to export or save the property tables to other programs like R or excel, so that I can calculate by myself and identify the relationships between properties?(I tried to find out how to do this but couldn't find the way so far..)

Thank you

I am conducting the study regarding to space syntax analysis. And as you know to do that, I need an axial map.

Actually I know the notion of axial map, and how to draw it in theoretically. As I know, I have to get an road outline map so that I can make an axial map, like the upper one that I uploaded. But in this country I found that I can get only this weird format just like the below.

I can't make all-line map and fewest-line map because these node poligons disrupt generating axial line. I mean when I press button that creates a all-line map, it only generates line between a node and a node.

I queried to the GIS platform website if there is an outline map that all roads are connected without that freaky node poligons, but they told me there's not. In this situation how the hell I can make an axial line map? The materials that teach how to draw an axial map only show a minimal space like a building and drawing it by hand in Autocad, not the large space that covers the certain region.

Of course I might draw the outline map on the AutoCAD, but the network that I want to analyze is an urban area that contains thousands of roads and I believe there might have been an easy way which I can perform it in software(like QGIS or depthmap). It won't be true that I have to draw all these by my hand.

Or do I have to use the road central line map? If so, please tell me what should I do with it?

Please help me. I'm gonna be crazy because of this...

Hello, I am a senior grade student in university and nowadays conducting a study about change of the urban structure(in terms of commercial activities) after certain transportation project.

After the project, several pedestrian roads will be installed(which is correspond to axial line in SSA) so that the number of people in the area will grow to much larger degree.

If I can quantify the differences of number of pedestrians between before/after of business commensement I will eventually reflect this as a variable to making a model(ex: regression, whatever) that analyzing commercial activity.

I've been studying SSM theory and notions, but couldn't find materials that I can get technical information...And as I heard, the procedure is AutoCAD or QGIS > depthmap > R, but so what? I have no idea how to do this in detail, and above all, I don't know how to use AutoCAD. So can anyone tell me how to analyize space syntax with QGIS? I've downloaded depthmapX 0.35 and QGIS space syntax toolkit already but don't know how to do further. I'm despairing now.

So please give your hand regarding this to this poor novice student :(

Of course i know it depends on the field and data i conducting the research in, and some fields like pharmacy or biology require very strict criteria.

But what i wonder is social science. I made a regression model to analyze the activity performed in urban space recently, which contains some variables whose p-value of t-test exceed 0.05 and rather close to 0.1. In this case can i apply relatively flexible criteria, 0.1 to each value?

And even if it's possible, applying criteria such as 0.15 won't be accepted, isn't it? Actually, several variables whose p-values are 0.1x are contained in the model

I would be very appreciate if you answer to my question and give a good reference regard to it.