Is it the reclass raster or the combination of raster IV factor and DV landslides?
Dear Edward,
The best way to calculate the logistic regression values of predictor variables outside ArcGIS ( can use SPSS or Prism or Statistica). The add that values in corresponding features and feature classes, use the raster calculator for the generation of final landslide susceptibility zonation map.
In ArcGIS also you can calculate the same using the tool box available for the regression analysis. But it is easier and more understandable, if you works outside ArcGIS for calculation of weights.
Vijith
you can develop a logistic regression in Weka software and then use the linear equation produced in raster calculator. this should be ok, \
Mr. Maher, do we need to separate the variables into 2 part (Continous & Categorical) during the analysis using SPSS for LR? And after we get the coefficient B..how to use the equation in the raster calculator? where to fit the coeff...in which raster? is it the reclass raster or how? thnks
Thanks Yannick, Actually my research is based on the HIgh Resolution LiDAR data to delineate the geohazard areas...I m using Frequency Ratio, Certainty Factor and Entropy model, but i need also to see the result if i were to use LR.? It involves about 65m points of 0.5m pixel.
Mr Yannick, actually I m extracting each point of all the factors/variables (IV) to landslide variable (DV)..all together is 65,000,000 point of 0.5m interval... i mean high resolution LiDAR data...
You may follow my Step by Step tutorial on Youtube channel "How to produce LSM using ArcGIS and R statistics" . If anything, please add a comment .
https://www.youtube.com/playlist?list=PL3GCqQd4xntAEkGUxx6c6Gi8XZBPtcQ9E
- Badges
- Science topic
- Similar topics
- Geoscience
- Cartography
- Mapping
- Maps