5

I have a shapefile which is located in province in indonesia. it contains many sub-province

if i removed sub regions inside a shapefile, and it leaves a small region, and my question is that small region's scale is as same as a previous scale into a map how to reset a new scale that i want?

PolyGeo
  • 65,136
  • 29
  • 109
  • 338
Q Shofwan Muhammad
  • 51
  • 2
  • 1
    Welcome to GIS SE! Our protocols can take some getting used to but the Q & A format works best if you ask just one question in each Question. I recommend that you edit your Question to focus it on the one that is the most important for you to get an Answer to first. Then ask the other two as separate Questions. – PolyGeo Nov 02 '13 at 08:03
  • Good advice in both answers. Because of 3 questions it is probably best to move to community wiki.
    But the best solution would be @q-shofwan-Muhammad move 2 questions to new question     – Brad Nesom Nov 02 '13 at 15:22
  • I've edited the question text and left only the question which has received an answer. Please post the other questions separately. – underdark Nov 03 '13 at 14:25

3 Answers3

4

That's really three questions, which breaks the Question/Answer model, but if you rewrite the question to only ask the third question the the answer could be:

All vector data is inherently georeferenced (only imagery can be unreferenced). It might not be in a known reference system (and therefore unusable with other data), but there's an origin and scaling factor in play. Exactly which coordinate reference system is being used can be inferred from a file with the same name as the shapefile, but an extension of .prj. It's not part of the shapefile specification, but is usually present. If it isn't present, then there are various brute force methods for determining what it might be (starting by looking at the range of values in the file), but that's a whole number of other questions.

There is no way to know the scale at which vector data was collected and processed, unless that information is passed as metadata (data about data). For shapefiles, Esri uses a .shp.xml suffix as the standard location (that's not part of the shapefile spec either). Without one there is no way of knowing the processing history of the data without speaking to every person who may have altered it (even with one, you still can't really know if some unlisted changes have occurred; from there depends on a chain of trust in good data hygiene).

Simple restriction (making a subset) does not change data in any way, and therefore does not change the scale of the data. It is only coordinate processing which can actually alter scale (usually only for the worse, not better). Data does not suddenly start showing details not previously present without supernatural involvement (most mapping systems use multiple datasets at different scales to handle this sort of magic).

Vince
  • 20,017
  • 15
  • 45
  • 64
  • 2
    +1 This is a good answer with good advice. The assertion in the second paragraph may be misunderstood, however, depending on how "inherently georeferenced" is understood. For instance, when I digitize a vector feature from an "unreferenced" image, your statement implies it somehow magically becomes georeferenced through that action by virtue of employing a vector format. Obviously that's impossible: there's a contradiction here. – whuber Nov 02 '13 at 13:32
  • Vectors have coordinates. I'd argue that even vectors digitized on an unreferenced image have a georeference, be it pixels, inches, or millimeters, oriented from some point along X and Y axis. Useful? No. But even vectors with no correlation to reality are referenced with respect to their origin. I suppose you could assert the same for unreferenced images, with respect to pixels, but I'd draw the line on integer resolution (and inflexible orientation). The key to making vector data useful is the use of a known reference system for the georeference. – Vince Nov 02 '13 at 17:18
  • In general terms a drawing that has coordinates, (all do), but is only referenced to to an origin (and possible scale). Are not geo-referenced. Geo-referencing is the process of applying a known projection (with origin, scale, rotation) which makes it usable with other geo-referenced data. Geo referenced data are data that know their geo location. Not to even mention distortions – Brad Nesom Nov 02 '13 at 17:46
3
  1. It is implied that shapefiles are geo referenced. However the data may be related to the real world in a couple of ways. Firstly, you should learn about Coordinate Reference Systems http://en.wikipedia.org/wiki/Spatial_reference_system (CRS). The CRS will have been decided by the creator of the shapefile and will ususally be included in the *.prj file included with the *.shp file (also will have a *.dbf and *shx).

Note that the *.prj file is plain text and should be able to be opened with common programs such as Notepad or Word.

QGIS creates it's own file with much the same type of information *.qpj

If there is no projection file included with the shapefile set then your problem is not uncommon What strategies, criteria, or rules to use for selecting coordinate systems?

For a guide related to qgis, slightly out of date, see

http://www.ga.gov.au/webtemp/image_cache/GA20953.pdf

You could try a crs that is used by convention by the authority that created the shapefile.

Community
  • 1
Willy
  • 3,500
  • 3
  • 24
  • 37
1

If no prj exists there is usually (by the software) implied wgs84 projection.
That would be coordinates like. 48.00001, -102.000001.

By scale I think you are actually asking about the extent (envelope).
To change that you will likely need to save the desired objects to a new file.
Esri has a method to recalculate this but I haven't ever seen it in qgis.
Maybe someone can correct this.

Brad Nesom
  • 17,412
  • 2
  • 42
  • 68
  • either that, or they are projected. Usually in a local CRS. You can tell a projected coordinate pair by looking at them. usually in that case x and y are in the thousands. – nickves Nov 03 '13 at 20:44