i can just use kriging by default and then use zonal stadistical as table for municipality level? (i have all the data from the stations on chile). I´m asking you because i was reading in some papers that i need to use a "global geo-referenced grid with 0.5°x0.5° resolution of monthly indicators" from climate research unit
Thank you, it was a great video with great explanation. I tried to implement the same method on my data, but couldn't plot the map. I got the error that my data is linear. Any idea?
I'm not good at stats, but in plain simple English (and not in wiki ultra-complex geometry), this sounds like creating a statistical probability deriving from the known or obtained mean values and upping it up by running through various filters or modes (linear, natural neighbor, etc.). The values of which are then ran through a 2D or 3D graph. Right? Much like standard deviation in stats, or in image resolution increase aka pixel upscaling (linear, bicubic, nearest neighbor)?
Thank you, can you share some method to make sure the parameters (for example, range,nugget ,lag distance, )of kriging rather than trying buy hand
i can just use kriging by default and then use zonal stadistical as table for municipality level? (i have all the data from the stations on chile). I´m asking you because i was reading in some papers that i need to use a "global geo-referenced grid with 0.5°x0.5° resolution of monthly indicators" from climate research unit
Thank you, it was a great video with great explanation. I tried to implement the same method on my data, but couldn't plot the map. I got the error that my data is linear. Any idea?
what does "R-squared" at 6:30 stand for ?
relative error
Thanks. Can you recommend some books?
You are an awsome lecturer...Go pls oxford to be a lecturer in university..ty a lot...perfect presenting
I'm not good at stats, but in plain simple English (and not in wiki ultra-complex geometry), this sounds like creating a statistical probability deriving from the known or obtained mean values and upping it up by running through various filters or modes (linear, natural neighbor, etc.). The values of which are then ran through a 2D or 3D graph. Right?
Much like standard deviation in stats, or in image resolution increase aka pixel upscaling (linear, bicubic, nearest neighbor)?
You guys don't support Mac do you?
50 / 5000
Please, could you send the data to practice
Very fancy video !