AWARES ExOb Wiki

User Tools

Site Tools

Trace: • idl
internal:administration:idl

Differences

This shows you the differences between two versions of the page.

Link to this comparison view

Both sides previous revisionPrevious revision
Next revision		Previous revision
internal:administration:idl [2025/07/22 15:26] – [execute shell command : SPAWN] janinternal:administration:idl [2026/04/28 08:26] (current) – [interpolation] jan
Line 8: Line 8:
  
 IDL current reference @ NV5 Geospatial software :\\ IDL current reference @ NV5 Geospatial software :\\
-[[https://www.nv5geospatialsoftware.com/docs/routines-1.html| IDL  8.9. Reference @ NV5 ]]\\ +[[https://www.nv5geospatialsoftware.com/docs/routines-1.html| IDL  Reference @ NV5 ]]\\ 
-[[https://www.nv5geospatialsoftware.com/docs/funclisting.html| IDL  8.9. funclisting @ NV5]]\\ +[[https://www.nv5geospatialsoftware.com/docs/funclisting.html| IDL  funclisting @ NV5]]\\ 
-[[https://www.nv5geospatialsoftware.com/docs/WhatsNew.html| IDL  8.9. whats new @ NV5]]\\+[[https://www.nv5geospatialsoftware.com/docs/WhatsNew.html| IDL  whats new @ NV5]]\\
  
  
Line 285: Line 285:
   IF x eq 4 THEN BEGIN   IF x eq 4 THEN BEGIN
     print, 'x equals 4' $     print, 'x equals 4' $
-  END ELSE BEGIN+  ENDIF ELSE BEGIN
     print, 'x is not 4'     print, 'x is not 4'
-  END+  ENDELSE 
 + 
 +to keep track of your blocks you may extend the END statements with the statement the block started with, i.e. ENDIF for the block started afer an 'IF' or ENDELSE or ENDCASE, ENDWHILE, etc. 
 +The interpreter tells you then if a ENDxxx statement does not fit. 
 + 
 +You may nest i.e. build structures like IF ... THEN ... ELSE IF ... THEN ...\\ 
 + 
 + 
 +Alternatively you use the CASE statement: 
 + 
 +  CASE x of 
 +     1: <> 
 +    22: BEGIN 
 +        ... 
 +        END 
 +    ELSE: BEGIN 
 +          ... 
 +          END 
 +  ENDCASE 
 + 
 +if you want to handle several cases of x in one block you can do this:
  
-for safety you may extend the END statements with blocked statement i.e. ENDIF or ENDELSE+  CASE 1 of 
 +     x eq 1: begin 
 +       ... 
 +        END 
 +    (x eq 22) OR (x eq 23): begin 
 +        ... 
 +        END 
 +    ELSE: begin 
 +          ... 
 +          end 
 +  ENDCASE
  
-you may nest i.e. build structures like IF ... THEN ... ELSE IF ... THEN ... 
-or use the CASE statement. 
  
 Loops can be implented with FOR, WHILE or REPEAT as one-liners: Loops can be implented with FOR, WHILE or REPEAT as one-liners:
Line 308: Line 336:
   FOR i = 0,5 DO BEGIN   FOR i = 0,5 DO BEGIN
     PRINT, i     PRINT, i
-    END+    ENDFOR
      
   i = 0   i = 0
Line 314: Line 342:
     i = i+1     i = i+1
     PRINT, i     PRINT, i
-    END+    ENDWHILE
      
   i = 1   i = 1
Line 320: Line 348:
     i = i + 1      i = i + 1 
     PRINT, i     PRINT, i
-  END UNTIL i gt 5+  ENDREP UNTIL i gt 5
  
 to make coding a bit safer you may append the name of the loop command to the END statement, i.e. to make coding a bit safer you may append the name of the loop command to the END statement, i.e.
Line 327: Line 355:
   ENDREP    ENDREP 
  
-a PROCEDURE is a subroutine that may take parameters and change them. Its definition is started with the word PRO and ends with a END statement+a PROCEDURE is a subroutine that may take parameters and change them. Its definition is started with the word PRO and ends with an END statement
  
   PRO my_proc, text_str, x_value=x_value   PRO my_proc, text_str, x_value=x_value
Line 598: Line 626:
   * As matrix operations are faster than for-loops this may speed up your code a lot ...   * As matrix operations are faster than for-loops this may speed up your code a lot ...
  
 +  * Alternatively one may use **rebin** , but note that rebin uses bilinear interpolation (sample=0) or nearest neigbor (sample=1) and accordingly might be a bit slower 
 +
 +  x = fltarr(Nx)
 +  y = fltarr(Ny)
 +  xx = rebin( x, Nx, Ny, sample=1 )
 +  yy = transpose(rebin( y, Ny, Nx ))
 +
 +  
 ==== procedure names ==== ==== procedure names ====
  
Line 787: Line 823:
   ss[0] = 'wind speed'   ss[0] = 'wind speed'
   ss[1] = 'm/s'   ss[1] = 'm/s'
 +
 +An array of strings can be also be splitted element by element, but only from Idl version 8. onwards. 
 +The result will be a  [[https://www.nv5geospatialsoftware.com/docs/list.html|LIST-Object]] , which might be converted to an array by using its method [[https://www.nv5geospatialsoftware.com/docs/list.html#ToArrayMethod|List::toArray()]] (note the brackets!).
 +
 +  s = [ 'a=b', 'c=d', 'e=f' ]
 +  sL = STRSPLIT( s, '=', /ex, count=N )
 +  sA = (STRSPLIT( s, '=', /ex, count=N )).toArray(  ) ; equal to sA = sL.toArray()
 +
 +Result %%sL%% will be a %%LIST%% object with N elements, and every entry of sL an array of strings with elements accorrding to the result fo strsplit.
 +Parameter %%N%% will be an array with the number of splits that could be made per element.
 +The conversion %%sL.toArray()%% works only if every entry in sL has the same number of elements.
 +The %%sA%% array would allow you to search for certain values in the array by using the where functions - eg:
 +
 +  ia = where( sA[*,0] eq 'a', Na )
 +
 +But you could also try to use the %%WHERE%% method of the %%LIST%% object ...
 +
  
  
Line 793: Line 846:
  
  
-filenames can be separated into path and name by +filenames can be separated into path and name with 
-  file_dirname( <path to file> ) +  f_dn = file_dirname( <path to file> ) 
-  file_basename( <path to file> )+  f_bn = file_basename( <path to file> ) 
 + 
 +where the dirname will have no '/' charcater. 
 + 
 +You may use strsplit to strip or get the 'filename extension' - the part of the filename after the last dot: 
 +  f_ext = (STRSPLIT( f_bn, '.', /ex, count=N ))[N-1] 
 +  f_bnx = STRJOIN((STRSPLIT( f_bn, '.', /ex, count=N ))[0:N-2],'.') 
  
 ==== formatted output ==== ==== formatted output ====
Line 1741: Line 1801:
     xrange = [ 0.5 * hour, 2*1e6 ], $     xrange = [ 0.5 * hour, 2*1e6 ], $
     /xlog, $     /xlog, $
-    xticks = N_ticks-1, $  ; holy IDL - why do you need things like this ?+    xticks = N_ticks-1, $  ; xticks is the number of //intervals// not ticks!,  holy IDL - why do you need things like this ?
     xtickname = x_tiknams, $     xtickname = x_tiknams, $
     xtickv    = x_tikvals, $     xtickv    = x_tikvals, $
Line 1845: Line 1905:
    1 => Plus sign (+)     1 => Plus sign (+) 
    2 => Asterisk (*)     2 => Asterisk (*) 
-   3 => Period (.) +   3 => one pixel dot (.) 
    4 => Diamond     4 => Diamond 
-   5 => Triangle 'up'+   5 => Triangle up (if symsize>0)  or down if symsize<0
    6 => Square     6 => Square 
    7 => X     7 => X 
Line 2505: Line 2565:
 ==== transparency ==== ==== transparency ====
  
-IDL does not know transparent overlays but on pixel based devices you can implement it with TVRD and TVRD.+IDL does not know transparent overlays but on pixel based devices you can implement it with TVRD and TV.
 Lets assume you want to put a bitmap BMP with size Nx, Ny at a position x0 y0 (pixels) with a tranparency factor of q_tra: Lets assume you want to put a bitmap BMP with size Nx, Ny at a position x0 y0 (pixels) with a tranparency factor of q_tra:
  
Line 2674: Line 2734:
 %%-dGraphicsAlphaBits=1%% avoids gridlines after rastering (http://www.idlcoyote.com/ps_tips/psstripes.html), \\ %%-dGraphicsAlphaBits=1%% avoids gridlines after rastering (http://www.idlcoyote.com/ps_tips/psstripes.html), \\
 %%-dEPSCrop%% crops the image at the bounding box. %%-dEPSCrop%% crops the image at the bounding box.
 +
 +You can execute gs from within IDL with the SPAWN command. As of IDL 9.x you need to prepend a ''unset LD_LIBRARY_PATH'' because IDL try to ensure its own (outdated) libraries for any spawn call:
 +  SPAWN, 'unset LD_LIBRARY_PATH; gs -dSAFER -dBATCH -dNOPAUSE -sDEVICE=png16m -r120 -dTextAlphaBits=4 -dGraphicsAlphaBits=1 -dEPSCrop -sOutputFile='+plt_name+'.png '+pl_name+'.eps', result, err_result
 +
 ==== postscript -> PDF ==== ==== postscript -> PDF ====
  
Line 3039: Line 3103:
   extrm_lev_max = +(machar()).xmax   extrm_lev_max = +(machar()).xmax
      
-See function [[https://www.harrisgeospatial.com/docs/machar.html|function machar]].+See function [[https://www.nv5geospatialsoftware.com/docs/MACHAR.html|function machar]].
  
  
Line 3055: Line 3119:
     /overplot     /overplot
  
-Keyword c_annotations is used here toprovide a format - this is usually not necessary.+Keyword c_annotations is used here to provide a format - this is usually not necessary.
 You can define also different thicknesses, colors etc. with the other %%c_...%% key words of %%contour%%  You can define also different thicknesses, colors etc. with the other %%c_...%% key words of %%contour%% 
 see [[http://www.physics.nyu.edu/grierlab/idl_html_help/C40.html#wp908085| help for contour]]. see [[http://www.physics.nyu.edu/grierlab/idl_html_help/C40.html#wp908085| help for contour]].
  
 +=== line patterns ===
 +
 +contour can draw hatch patterns - but note: the pattern is not transparent.
 +
 +  contour, z, x, y, $
 +    levels = levels, $
 +    c_orientation = [ 0 , 22, 45, 90, ... ] , $ ; orientation of the lines in degrees
 +    c_spacing = [ 1, 0.5, 0.25, ... ] ; spacing between lines in cm
 +    
 +
 +=== Path of contour lines ===
 +
 +contour can provide the pathes of the contour lines:
 +
 +  contour, z, x, y, $
 +    levels = levels, $ 
 +    path_info = path_info, $ ; info about the provided pathes (see below). If you provide here a variable contour will not draw.
 +    path_xy = path_xy, $ ; variable for coordinates: fltarr(2,N_tot)
 +    /path_data_coords, $ ; want to have data coordinates in data_xy instead of normalized coo's.
 +    /path_double, $ ; you need double if eg x is time as julian day
 +    /closed, $ ; close pathes along outline of the plotting area when isolines leave it.
 +    /overplot ; avoid contour to rescale the axes - if you already had a plot or contour plot you want to reuse
 +
 +path_info is an array of CONTOUR_PATH_STRUCTURE. The number of elelemtns gives the number of sub pathes.
 +Elements %%offset%% and %%N%% give the starting index and length of every isoline in path_xy.
 +
 +You can then use the pathes in poly_fill to do transparent hatch patterns: 
 +
 +  for i = 0, n_elements(path_info)-1 do begin
 +    ii = path_info[i].offset + lindgen(path_info[i].N)
 +    polyfill, path_xy[0,ii], path_xy[1,ii], /line_fill, spacing=0.7, orientation=i*5, noclip=0
 +  endfor ; i
 +
 +You could also use polyfill to put a byte pattern or wrap an image between the isolines ...
 +
 +  c = 255
 +  pattern = [ $
 +      [ 0, 0, 0, 0, 0, 0, 0 ], $
 +      [ 0, 0, 0, c, 0, 0, 0 ], $
 +      [ 0, 0, c, c, c, 0, 0 ], $
 +      [ 0, c, c, c, c, c, 0 ], $
 +      [ 0, 0, c, c, c, 0, 0 ], $
 +      [ 0, 0, 0, c, 0, 0, 0 ], $
 +      [ 0, 0, 0, 0, 0, 0, 0 ]  $
 +    ]
 +  for i = 0, n_elements(path_info)-1 do begin
 +    ii = path_info[i].offset + lindgen(path_info[i].N)
 +    polyfill, path_xy[0,ii], path_xy[1,ii], pattern=pattern, transparent=c-1, noclip=0, color='ff00ff'x
 +  endfor ; i
 +
 +But i do not understand how the transparent thing works - in this example it does not work. And if i understood it right i need to provide IMAGE_COORD, 
 +see  NV5 help for [[https://www.nv5geospatialsoftware.com/docs/POLYFILL_Procedure.html#dg_routines_3604229493_877795|POLYFILL]].
  
 === color bar === === color bar ===
  
-You can use contour also to make a **color bar**:+You may want to have a color bar setting the colorshading in relation to values. 
 +You can use contour to make a **color bar**:
  
 A **horizontal** color bar at the bottom A **horizontal** color bar at the bottom
Line 3244: Line 3361:
 A one dimensional histogram is e.g. generated with A one dimensional histogram is e.g. generated with
  
-  histo = histogram( data , binsize=bin, locations = classes , /nan )+  binsize = 0.1 
 +  histo = histogram( data , binsize=binsize, locations = classes , /nan )
  
 the classes variable will contain the **lower borders** of the bins with the first bin starting at min(data). \\ the classes variable will contain the **lower borders** of the bins with the first bin starting at min(data). \\
Line 3263: Line 3381:
  
 or you write an own plot-histo procedure ... or you ask jan :-) ... or you write an own plot-histo procedure ... or you ask jan :-) ...
 +
 +
 +A histogram with geometric spacing, i.e. each bin is by a factor q_bin larger than the previous.\\
 +In this case data points lower or equal to zero are not allowed:
 +  q_bin = 2.0
 +  log_binsize = alog10(q_bin)
 +  histo = histogram( alog10(data[where(data gt 0)]), binsize=log_binsize, locations = log_bins, /nan )
 +  classes = 10^log_bins
 +
 +Plotting shall use then logarithmic scaling for the x-axes:
 +
 +    plot, classes , histo , psym=10, $
 +      xtitle = 'class (unit)', $
 +      ytitle = 'freq.', $
 +      /xlog
 +
  
 === 2-D histogram === === 2-D histogram ===
Line 3273: Line 3407:
   max_data_1 = max(data_1, /nan )   max_data_1 = max(data_1, /nan )
   N_bins_1 = fix((max_data_1-min_data_1)/bin_size_1)+1   N_bins_1 = fix((max_data_1-min_data_1)/bin_size_1)+1
-  bins_1 = min_data_1 + (findgen(N_bins_1) + 0.5) * bin_size_1 ; output of histo refers to lower border of bins => add 1/2binsize to center them in bin+  ; output of histo refers to lower border of bins => add 1/2binsize to center them in bin 
 +  bins_1 = min_data_1 + (findgen(N_bins_1) + 0.5) * bin_size_1 
        
   ; set bin-size, min and max and vector with bin-borders for data set 2   ; set bin-size, min and max and vector with bin-borders for data set 2
Line 3280: Line 3415:
   max_data_2 = max(data_2, /nan )   max_data_2 = max(data_2, /nan )
   N_bins_2 = fix((max_data_2-min_data_2)/bin_size_2)+1   N_bins_2 = fix((max_data_2-min_data_2)/bin_size_2)+1
 +  ; output of histo refers to lower border of bins => add 1/2binsize to center them in bin
   bins_2 = min_data_2 + (findgen(N_bins_2) + 0.5)*bin_size_2   bins_2 = min_data_2 + (findgen(N_bins_2) + 0.5)*bin_size_2
        
Line 3715: Line 3851:
 If you want to interpolate only over certain dimensions they must be the last ones. You may use transpose to rearrange. Lets assume you have an array depending on x, y and time, i.e. z=fltarr(Nx,Ny,Nt) and you want to interpolate to certain x and y coordinates you can do this as follows: If you want to interpolate only over certain dimensions they must be the last ones. You may use transpose to rearrange. Lets assume you have an array depending on x, y and time, i.e. z=fltarr(Nx,Ny,Nt) and you want to interpolate to certain x and y coordinates you can do this as follows:
  
-  z_ipol = transpose( interpolate( transpose(z_org,[2,0,1]) , ix_ipol , iy_ipol , /grid  ), [1,2,0] )+  z_ipol = transpose( interpolate( transpose(z_org,[2,0,1]) , ix_ipol , iy_ipol , /grid, MISSING=!VALUES.f_nan  ), [1,2,0] )
  
 +The /GRID flag allow to use different sizes for ix_ipol and iy_ipol and the MISSING keyword sets values of z_ipol outside the ranges of z_org to the missing value (no extrapolation).
  
  
internal/administration/idl.1753198010.txt.gz · Last modified: by jan