Reputation
47,011
Next tag badge:
150/100 score
19/20 answers
Badges
9 118 187
Newest
 Enlightened
Impact
~1.6m people reached

1d
comment Rectangle shapefiles with the same extents does not align perfectly after projection
It depends on how the shapefiles were created! If one was digitized in a particular projection, then when displayed in that projection it will be as accurate as the digitization process could make it, regardless of how coarsely segmented it might be. This is an issue that ultimately can be decided only by referring to metadata and having a clear understanding of what you are trying to accomplish and what changes might be made to your data during the process.
1d
comment Calculate visibility with polygonal visual barriers
The problem is that this does not "exclude" the forest areas. On the contrary, your approach causes them to block all lines of sight that pass through any forested area. It seems unlikely that such behavior would be intended in any application.
1d
comment Formally testing whether building type tends to be close to another building type?
@Andre The literature suggests that as few as 200 permutations should be enough to compute a p-value. Since most of the work goes into computing the distance matrix, one might as well use many more. The code in this answer by default uses 1e3--that is, 1000--mainly because it takes little time. In publications I have seen people use between 50 and 100,000 simulations for work like this. (50 were used when each permutation was very computationally expensive.) I'm not sure what distinction you are making between "sample size" and "number of simulations," though.
1d
comment Calculate visibility with polygonal visual barriers
This multiplication will be ineffective for two reasons. First, at locations away from sea level your forested areas will rise many kilometers into the sky, which is obviously unrealistic. Second, near sea level they will be too short. Why not just add a typical tree height to the DEM in forested areas?
1d
comment Tool to determine the trend of data point positions?
Could you define, explain, or otherwise quantify exactly what you mean by "trend"?
2d
comment Assigning points to grid using R
I am sorry for the poor reception your post got here, Chris. IMHO it's perfectly clear and (regardless of any issues about clarity) the downvotes are undeserved. They reflect poorly on our site, not on you. To help you out, I have modified my answer in the post you link to so that it will easily cope with arbitrary grids: simply specify a cellsize of c(0.25,0.25) in the ji function.
2d
revised Aggregating points to grid using R
added 349 characters in body
2d
comment Calculate visibility with polygonal visual barriers
Your idea of modifying the elevations is an effective one. (It can even be extended to allow for random variation in heights within each environment and even partial transparency if you wish.) For how to make the modifications, search our site for con.
Apr
27
reviewed Leave Open NTv2 inverse transformation calculation process
Apr
25
comment Distance to river square grids
The first question is answered in many posts here. The second question is unintelligible: what exactly is your problem, what is a "certain feature," and what do you mean by "tips"?
Apr
25
comment Are perpendicular lines still perpendicular after converting from UTM zone 15 to 16 (or between two neighboring zones)?
The important thing here is to recognize that these changes of projection will change the lengths of the sides of figures. They will also tend to change them disproportionately: almost all the change occurs in the east-west direction. As Matte points out, those length changes for a figure straddling two UTM zones will be tiny. That gets back to the more fundamental point: since they're tiny, why bother?
Apr
23
comment Are perpendicular lines still perpendicular after converting from UTM zone 15 to 16 (or between two neighboring zones)?
A "square," by definition, has equal sides as well as 90 degree angles. Figures with four (geodesic) sides and four 90 degree angles are called rectangles. Although these don't actually exist on the sphere, close approximations of them do, so it's worthwhile respecting these definitions.
Apr
23
answered Formally testing whether building type tends to be close to another building type?
Apr
23
comment Are perpendicular lines still perpendicular after converting from UTM zone 15 to 16 (or between two neighboring zones)?
This argument is irrelevant, because there's still little distortion at UTM zone boundaries. In fact, UTM zones are designed to work outside their margins. I forget the size of the allowance, but it's around 1/2 to 1 degree at the Equator. At higher latitudes one can move even further east or west before encountering any distortion that's greater than the maximum encountered within any single UTM zone.
Apr
23
comment Are perpendicular lines still perpendicular after converting from UTM zone 15 to 16 (or between two neighboring zones)?
The square will not remain square--but the angles will remain at 90 degrees, according to the (correct) argument in the answers by csd and nickves.
Apr
23
comment unexpected output from raster calculator function
Sam's formula is the correct one for what he wants it to do. See en.wikipedia.org/wiki/Logit.
Apr
23
comment unexpected output from raster calculator function
@Nir That's simply incorrect. Sam: Could you explain how you know all the values are zero? Sometimes ArcMap's automatic legend is screwed up but the values in the dataset are still ok.
Apr
21
comment Lambert conformal conic projection
I had not critically evaluated such pictures before now. (Snyder doesn't include one.) Most of those I can find look bogus, because they do not show specifically how the sphere is projected onto the cone. How the scale varies is not in the least apparent from the formulas, either. In particular, it sure looks like features have (eventually) to be shrunk to fit near the apex of the cone! Here's a reasonably accurate picture: mathworld.wolfram.com/ConicProjection.html . It shows that there is a unique parallel away from which scale increases.
Apr
21
comment Lambert conformal conic projection
This answer fails to explain the specific phenomenon called out in the question; namely, why does scale decrease between the parallels and increase outside them?
Apr
21
comment problem with raster calculator syntax
Regardless of the software, there are obvious typographical errors. For instance, "11.25" has been concatenated to a string expression--no raster calculator will know what to make of that. The first thing to do is fix them. Even if you're successful, it would be wise to compute each sub-expression separately and review its results to check that you are getting what you expect. That exercise will pinpoint any remaining problems.