Thank you so much for this great class. When I select a few observations in the scatter plot matrix, I don't get any new lines like you do. The ones i selected are highlighted but the only line is the one that was there before. Is there a setting where I can fix this?
I'm not sure- I don't recall changing any settings, since I don't use GeoDa that much. I hope you can get it figured out! You might try getting help here: geodacenter.github.io/support.html
Hello, thank you for the video. I am working on risk factors of malnutrition in Uganda. But the challenge I have is getting the UDHS data into Geoda. Whenever I try, I get a message that "Failed to open data source" and thus cant proceed. What could be the problem.
Thank you for your very clear tutorial. Anyway, I have a very basic question about this spatial analysis in Geoda, so can we do a time series spatial analysis? for example, if I have a GDP data from 2010 to 2017, can I make a simultaneous estimation through those years? or should I do it separately year by year? and then compare it. thanks in advance.
You could do very basic spatial time series analysis in GeoDa, but nothing fancy.Honestly, I am not a time series person at all (and certainly not a Macroeconomist), so my advice is probably not the best. Analyzing GDP (from the little I understand) properly takes very complicated, specialized models that GeoDa can't handle. An extremely simple method would be to create a dummy variable for each year (except one) and allow those variables to control for business cycles. But, to do it properly takes specialized time series training that I don't have. Best of luck!
Thank you for the video. I'm working on my project with time-series cross-sectional data using GeoDa, have been using time editor also & made it into group varibles, but i got confused when i wanna regressed it the variable came by each sepatared year. Does Geoda can't run it like SPSS? Pardon me for my scramble english.
GeoDa does not have as many options for estimating complicated models. For that, you will need to use Matlab, or possibly R, although Stata and SAS have decent support for spatial models now.
this is so amazing. we can see the spread of a case (eg disease or crime) based on the number of events in county levels. i wonder how about if we use the points data. Is it applicable ?
Yes, I think a lot of people use points data, though I usualy do not. Here is a random link discussing such a case web.sgh.waw.pl/~atoroj/ekonometria.../4_pure_SAR_GIS_points_grids_EN.pdf