I have a shapefile (or several shapefiles, one per year, that I was going to join together to get all of the data in one attribute table) with an attribute for surface water area for each year. So like this: area_2000, area_2001, ..., area_2015; each year with about 2 million objects/rows.

I was hoping to do a linear trend time series analysis in ArcGIS Desktop to get the slope, p-value, etc. as additional attributes for each object. I have used ArcGIS Desktop before to do regression analysis, but this time I want to do time series to get the trend in area for each object.

Is there a tool that would support this?

  • 1
    How is this different to regression analysis? – FelixIP Jul 24 '18 at 23:56
  • As I understand it the regression analysis in Arc looks at how multiple factors (represented by shapefiles) might predict a dependent variable (also another shapefile layer). I am just looking to get a trend across time. Maybe I can use the standard regression analysis tools but I'm having trouble wrapping my head around how I could do that. – user20408 Jul 25 '18 at 1:03
  • 1
    ESRI dos support this type of analysis. Also, please provide more information on your desired analysis. A trend analysis is different than a multivariate temporal model. What is your dependent variable and what distributional form does it take? – Jeffrey Evans Jul 25 '18 at 1:15
  • Sorry; "does not support" – Jeffrey Evans Jul 25 '18 at 1:55

This question has no GIS component, but anyway. I ran this field calculator expression:

import numpy as np
def getSlope(y):
 a,b=np.polyfit(x, y, 1)
 return a
getSlope([ !Y_2001!, !Y_2002!, !Y_2003!, !Y_2004!, !Y_2005!])

on the field "SLOPE" in below table:

enter image description here

To get a slope of a trend. Look at numpy functions closely.

| improve this answer | |
  • Thank you Felix. I think I will use this as inspiration and do the math in arcpy. I had just wanted to see if there was a built in Arc function to do this before I tried to write it in arcpy. – user20408 Jul 25 '18 at 15:49

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.