list values in a table or array???

[irinivozinaki]

Hello everyone!

I would like to ask you if there is a way in python to do the following:

I have 20 values and I want to list them in a 5x4 array. Or is it better to list them in a 5x4 table? Which is the difference between the table and array in python???

(Is this helpful?) I need to save this table/array as an ascii file, which will be continuously converted in a raster file!




Thank you very much in advance!!!

[ArkadiuszMatoszka]

Hi,
In my considered opinion there is no reason to use either of this to for this purpose. Simple list slicing with defined number of columns will do the trick.

Simple example:

i = range(20)
cols = 4
f = open('D:\\tmp.csv', 'w')
for r in range(len(i)/cols):
    f.write(';'.join([str(v) for v in i[r*cols:(r+1)*cols]]) + '\n')

Regards
Arek

[RaphaelR]

Hi Irini,

for Arrays in python numpy is really handy.
There are also arcpy functions to convert numpy Arrays to rasters and the other way around.
http://help.arcgis.com/en/arcgisdesktop/10.0/help/index.html#//000v00000130000000

for example

import numpy as np

a = [1,2,3,4,5,6,7,8,9,10,11,12,13,14,15,16,17,18,19,20]

a = np.array(a)
a = np.reshape(a, (4,5))

will get you:

[[ 1  2  3  4  5]
 [ 6  7  8  9 10]
 [11 12 13 14 15]
 [16 17 18 19 20]]

[ArkadiuszMatoszka]

There are also arcpy functions to convert numpy Arrays to rasters and the other way around.
http://help.arcgis.com/en/arcgisdesktop/10.0/help/index.html#//000v00000130000000

Didn't know about that. Always something new to learn 🙂
Cheers.
A

[zhengniu]

it is depending on how you will use this table or array, what they are used for.

[irinivozinaki]

Thank you all very much!!!

More accurately,

I have created part of a code which

a) takes two specific raster files (of a 5x4 grid) following a path

b) creates two separate arrays of 5x4 dimension
    I have used the following:
    array_d=arcpy.RasterToNumPyArray(raster_d)
    array_v=arcpy.RasterToNumPyArray(raster_v)
    (height_d, width_d)=array_d.shape
    (height_v, width_v)=array_velocity.shape

c) Subsequently, the code reads every cell value of both arrays, obtaining two values h, v

    for row1 in range (0,height_d):

        row2=row1

        for col1 in range (0,width_d):

            col2=col1
      
            h=array_d.item(row1,col1)
            v=array_v.item(row2,col2)

d) according to several scenarios of these values, h and v, the code calculates a parameter, named "D", in every cell. That is, the D parameter depends on the h and v.


I need to save the calculated parameter, "D", for every cell in a new array of the same dimensions (5x4 dimension), in an ascii file. 
This ascii file will be converted subsequently to a raster file (I suppose this is the easy thing)

[RaphaelR]

not sure if i understand correctly but if you´re working on cell by cell basis, wouldn´t this be much easier with just using raster calculator?
anyways, code below creates a new array for the D parameter and writes it to an ascii file (no header, will need one if you want to use AsciiToRaster).

array_d=arcpy.RasterToNumPyArray(raster_d)
array_v=arcpy.RasterToNumPyArray(raster_v)
(height_d, width_d)=array_d.shape
(height_v, width_v)=array_v.shape

# create empty array to store the "D" parameter
newarr = np.zeros((5,4))


for row1 in range (0,height_d):

    row2=row1

    for col1 in range (0,width_d):
    
        col2=col1

        h=array_d.item(row1,col1)
        v=array_v.item(row2,col2)
        
        # calculate the "D" parameter
        expr = h+v
        
        newarr[row1,col1] =  expr

# write the new array to an ascii file
np.savetxt("e:/out.txt", newarr, fmt="%.2f", delimiter= " ")

[irinivozinaki]

@ rafaelr 

Your help was very important! My problem is solved for small grids! Let's try for bigger ones!!!

I am not good on python programming but I am trying...

I would like to ask you for the raster calculator.
As I can see it is an ArcGIS tool, but as long as the calculations among my cells are not the same every time, for all the cells.
For example I am checking whether the cell value of the first array is greater than one and the cell value of the second array is lower than 3 and if this happens the D value is calculated through a specific equation. But there are several scenarios for the read cell values...

So...It won't help me, will it?

Thank you again!!!

[RaphaelR]

you can build elaborate expressions with raster calculator as well:
http://help.arcgis.com/en/arcgisdesktop/10.0/help/index.html#//009z000000z7000000.htm

however, going with python only is sometimes more straight forward (long/complex expressions in raster calculator can become a bit hard to read). i guess it´s a preference thing.

[irinivozinaki]

Yesterday I supposed that converting the ascii to raster file would be the easy thing. May be I was wrong 😞

Something goes wrong with the needed header of the ascii file.

I have created an ascii file (damageFile) which is the same with the one where the calculated values are saved.

I am writing the following...

damageFile=open('C:\\Python26\\ArcGIS10.0\\Python programming\\test2.ascii', "w+")
damageFile.write('NCOLS 5' + "\n")
damageFile.write('NROWS 4' + "\n")
damageFile.write('XLLCORNER 513589.931' + "\n")
damageFile.write('YLLCORNER 3923728.436' + "\n")
damageFile.write('CELLSIZE 10' + "\n")
damageFile.write('NODATA_VALUE -100' + "\n")

then the calculations follow under several scenarios...

and then the statement:

numpy.savetxt("C:\\Python26\\ArcGIS10.0\\Python programming\\test2.ascii", Damage, fmt="%14.11f", delimiter="w+")

and the damageFile is closed...


Can you find any mistake that I can see???

Thank you very much!!!