|
1 ;---------------------------------------------------------------------- |
|
2 ; add Month to output |
|
3 ; add lat, lon to output |
|
4 ; add year to output |
|
5 ; |
|
6 ; This example reads an ASCII file that is formatted a specific way, and |
|
7 ; writes out the results to a netCDF file. |
|
8 ; |
|
9 ; The first line in the ASCII file must be a header, with each field |
|
10 ; separated by a single character delimiter (like a ","). The rest of |
|
11 ; the file must be such that each row contains all fields, each |
|
12 ; separated by the designated delimiter. |
|
13 ; |
|
14 ; The fields can be integer, float, double, character, or string. |
|
15 ; String fields cannot be written to a netCDF file. They have to |
|
16 ; be read in as character arrays and written out that way. |
|
17 ;---------------------------------------------------------------------- |
|
18 |
|
19 ;---------------------------------------------------------------------- |
|
20 ; This function returns the index locations of the given delimiter |
|
21 ; in a row or several rows of strings. |
|
22 ;---------------------------------------------------------------------- |
|
23 function delim_indices(strings,nfields,delimiter) |
|
24 local cstrings, cdelim |
|
25 begin |
|
26 nrows = dimsizes(strings) |
|
27 ; |
|
28 ; Handle special case if we only have one string. Make sure it |
|
29 ; is put into a 2D array. |
|
30 ; |
|
31 if(nrows.eq.1) then |
|
32 cstrings = new((/1,strlen(strings)+1/),character) |
|
33 end if |
|
34 |
|
35 cstrings = stringtochar(strings) ; Convert to characters. |
|
36 cdelim = stringtochar(delimiter) ; Convert delimiter to character. |
|
37 ; |
|
38 ; Som error checking here. Make sure delimiter is one character. |
|
39 ; |
|
40 nc = dimsizes(cdelim) |
|
41 rank = dimsizes(nc) |
|
42 if(rank.ne.1.or.(rank.eq.1.and.nc.ne.2)) then |
|
43 print("delim_indices: fatal: the delimiter you've selected") |
|
44 print("must be a single character. Can't continue.") |
|
45 exit |
|
46 end if |
|
47 |
|
48 ; |
|
49 ; Create array to hold indices of delimiter locations, and then loop |
|
50 ; through each row and find all the delimiters. Make sure each row has |
|
51 ; the correct number of delimiters. |
|
52 ; |
|
53 ndelims = nfields-1 |
|
54 cindices = new((/nrows,ndelims/),integer) |
|
55 do i = 0, nrows-1 |
|
56 ii = ind(cstrings(i,:).eq.cdelim(0)) |
|
57 ; |
|
58 ; Make sure there were delimiters on this row. If not, we just quit. |
|
59 ; This could probably be modified to do this more gracefully. |
|
60 ; |
|
61 if(any(ismissing(ii))) then |
|
62 print("delim_indices: fatal: I didn't find any delimiters") |
|
63 print("('" + delimiter + "') on row " + i + ". Can't continue.") |
|
64 exit |
|
65 end if |
|
66 if(dimsizes(ii).ne.ndelims) then |
|
67 print("delim_indices: fatal: I expected to find " + ndelims) |
|
68 print("delimiters on row " + i + ". Instead, I found " + dimsizes(ii) + ".") |
|
69 print("Can't continue.") |
|
70 exit |
|
71 end if |
|
72 |
|
73 cindices(i,:) = ii |
|
74 |
|
75 delete(ii) ; For next time through loop |
|
76 end do |
|
77 |
|
78 return(cindices) |
|
79 end |
|
80 |
|
81 ;---------------------------------------------------------------------- |
|
82 ; This function reads in a particular field from a string array, |
|
83 ; given the field number to read (fields start at #1 and go to #nfield), |
|
84 ; and the indices of the delimiters. |
|
85 ; |
|
86 ; It returns either an integer, float, double, character, or a string, |
|
87 ; depending on the input flag "return_type". |
|
88 ;---------------------------------------------------------------------- |
|
89 function read_field(strings,ifield,indices,return_type) |
|
90 local nstring, cstrings, nf, tmp_str |
|
91 begin |
|
92 nrows = dimsizes(strings) |
|
93 ; |
|
94 ; Handle special case if we only have one string. Make sure it |
|
95 ; is put into a 2D array. |
|
96 ; |
|
97 if(nrows.eq.1) then |
|
98 cstrings = new((/1,strlen(strings)+1/),character) |
|
99 end if |
|
100 |
|
101 cstrings = stringtochar(strings) |
|
102 nf = dimsizes(indices(0,:))+1 ; indices is nrows x (nfields-1) |
|
103 |
|
104 ; |
|
105 ; Error checking. Make sure user has entered a valid field. |
|
106 ; |
|
107 if(ifield.le.0.or.ifield.gt.nf) then |
|
108 print("read_field: fatal: you've selected a field that is") |
|
109 print("out-of-range of the number of fields that you have (" + nf + ").") |
|
110 exit |
|
111 end if |
|
112 |
|
113 ; |
|
114 ; Set up array to return. For string, int, float, or double arrays, |
|
115 ; we don't have to do anything special. For character arrays, |
|
116 ; however, we do. |
|
117 ; |
|
118 if(return_type.ne."character") then |
|
119 return_array = new(nrows,return_type) |
|
120 else |
|
121 ; |
|
122 ; We don't know what the biggest character array is at this point, so |
|
123 ; make it bigger than necessary, and then resize later as necessary. |
|
124 ; |
|
125 tmp_return_array = new((/nrows,dimsizes(cstrings(0,:))/),"character") |
|
126 |
|
127 max_len = 0 ; Use to keep track of max lengths of strings. |
|
128 end if |
|
129 |
|
130 do i = 0,nrows-1 |
|
131 ; |
|
132 ; Special case of first field in row. |
|
133 ; |
|
134 if(ifield.eq.1) then |
|
135 ibeg = 0 |
|
136 iend = indices(i,ifield-1)-1 |
|
137 else |
|
138 ; |
|
139 ; Special case of first field in row. |
|
140 ; |
|
141 if(ifield.eq.nf) then |
|
142 ibeg = indices(i,ifield-2)+1 |
|
143 iend = dimsizes(cstrings(i,:))-1 |
|
144 ; |
|
145 ; Any field between first and last field. |
|
146 ; |
|
147 else |
|
148 ibeg = indices(i,ifield-2)+1 |
|
149 iend = indices(i,ifield-1)-1 |
|
150 end if |
|
151 end if |
|
152 ; |
|
153 ; Here's the code that pulls off the correct string, and converts it |
|
154 ; to float if desired. |
|
155 ; |
|
156 if(return_type.eq."integer") then |
|
157 return_array(i) = stringtointeger(chartostring(cstrings(i,ibeg:iend))) |
|
158 end if |
|
159 if(return_type.eq."float") then |
|
160 return_array(i) = stringtofloat(chartostring(cstrings(i,ibeg:iend))) |
|
161 end if |
|
162 if(return_type.eq."double") then |
|
163 return_array(i) = stringtodouble(chartostring(cstrings(i,ibeg:iend))) |
|
164 end if |
|
165 if(return_type.eq."string") then |
|
166 return_array(i) = chartostring(cstrings(i,ibeg:iend)) |
|
167 end if |
|
168 if(return_type.eq."character") then |
|
169 if( (iend-ibeg+1) .gt. max_len) then |
|
170 max_len = iend-ibeg+1 |
|
171 end if |
|
172 tmp_return_array(i,0:iend-ibeg) = cstrings(i,ibeg:iend) |
|
173 end if |
|
174 end do |
|
175 |
|
176 if(return_type.eq."character") then |
|
177 return_array = new((/nrows,max_len/),"character") |
|
178 return_array = tmp_return_array(:,0:max_len-1) |
|
179 end if |
|
180 |
|
181 return(return_array) |
|
182 end |
|
183 |
|
184 |
|
185 ;---------------------------------------------------------------------- |
|
186 ; This function reads in string fields only to get the maximum string |
|
187 ; length. |
|
188 ;---------------------------------------------------------------------- |
|
189 function get_maxlen(strings,ifield,indices) |
|
190 local nstring, cstrings, nf, tmp_str |
|
191 begin |
|
192 nrows = dimsizes(strings) |
|
193 ; |
|
194 ; Handle special case if we only have one string. Make sure it |
|
195 ; is put into a 2D array. |
|
196 ; |
|
197 if(nrows.eq.1) then |
|
198 cstrings = new((/1,strlen(strings)+1/),character) |
|
199 end if |
|
200 |
|
201 cstrings = stringtochar(strings) |
|
202 nf = dimsizes(indices(0,:))+1 ; indices is nrows x (nfields-1) |
|
203 |
|
204 ; |
|
205 ; Error checking. Make sure user has entered a valid field. |
|
206 ; |
|
207 if(ifield.le.0.or.ifield.gt.nf) then |
|
208 print("read_field: fatal: you've selected a field that is") |
|
209 print("out-of-range of the number of fields that you have (" + nf + ").") |
|
210 exit |
|
211 end if |
|
212 ; |
|
213 ; We don't know what the biggest character array is at this point, so |
|
214 ; make it bigger than necessary, and then resize later as necessary. |
|
215 ; |
|
216 tmp_return_array = new((/nrows,dimsizes(cstrings(0,:))/),"character") |
|
217 |
|
218 max_len = 0 ; Use to keep track of max lengths of strings. |
|
219 |
|
220 do i = 0,nrows-1 |
|
221 ; |
|
222 ; Special case of first field in row. |
|
223 ; |
|
224 if(ifield.eq.1) then |
|
225 ibeg = 0 |
|
226 iend = indices(i,ifield-1)-1 |
|
227 else |
|
228 ; |
|
229 ; Special case of first field in row. |
|
230 ; |
|
231 if(ifield.eq.nf) then |
|
232 ibeg = indices(i,ifield-2)+1 |
|
233 iend = dimsizes(cstrings(i,:))-1 |
|
234 ; |
|
235 ; Any field between first and last field. |
|
236 ; |
|
237 else |
|
238 ibeg = indices(i,ifield-2)+1 |
|
239 iend = indices(i,ifield-1)-1 |
|
240 end if |
|
241 end if |
|
242 if( (iend-ibeg+1) .gt. max_len) then |
|
243 max_len = iend-ibeg+1 |
|
244 end if |
|
245 end do |
|
246 |
|
247 return(max_len) |
|
248 end |
|
249 |
|
250 ;---------------------------------------------------------------------- |
|
251 ; Main code. |
|
252 ;---------------------------------------------------------------------- |
|
253 begin |
|
254 |
|
255 ; Set up defaults here. We are hard-coding the field types here. |
|
256 ; You can set up this script to try to determine the field types |
|
257 ; automatically, but this is a bit tedious. Maybe later. |
|
258 |
|
259 station = "USHa1" |
|
260 lat = 42.5378 |
|
261 lon = -72.1715 + 360. |
|
262 year = 1991 |
|
263 nfields = 30 ; # of fields |
|
264 delimiter = "," ; field delimiter |
|
265 |
|
266 filename = station+year+"_L4_m.txt" ; ASCII" file to read. |
|
267 cdf_file = station+year+"_L4_m.nc" ; netCDF file to write. |
|
268 |
|
269 ; In this case, fields #2-#2 are integers, |
|
270 ; and the rest of the fields are floats. |
|
271 |
|
272 var_types = new(nfields,string) |
|
273 var_strlens = new(nfields,integer) ; var to hold strlens, just in case. |
|
274 |
|
275 var_types = "float" ; Most are floats. |
|
276 var_types(0:1) = "integer" |
|
277 |
|
278 if(isfilepresent(cdf_file)) |
|
279 print("Warning: '" + cdf_file + "' exists. Will remove it.") |
|
280 system("/bin/rm " + cdf_file) |
|
281 end if |
|
282 |
|
283 ; Read in data as strings. This will create a string array that has the |
|
284 ; same number of strings as there are rows in the file. We will then need |
|
285 ; to parse each string later. |
|
286 |
|
287 read_data = asciiread(filename,-1,"string") |
|
288 header = read_data(0) ; Header. Use for variable names. |
|
289 data = read_data(1:) ; Get rid of first line which is a header. |
|
290 nrows = dimsizes(data) ; Number of rows. |
|
291 |
|
292 ; Read in locations of delimiters in each string row. |
|
293 |
|
294 hindices = delim_indices(header,nfields,delimiter) ; header row |
|
295 dindices = delim_indices(data,nfields,delimiter) ; rest of file |
|
296 |
|
297 ; print (hindices) |
|
298 ; print (dindices) |
|
299 |
|
300 ; Read in the field names which will become variable names on |
|
301 ; the netCDF file. |
|
302 |
|
303 var_names = new(nfields,string) |
|
304 |
|
305 do i=0,nfields-1 |
|
306 var_names(i) = read_field(header,i+1,hindices,"string") |
|
307 end do |
|
308 |
|
309 ; Write out this netCDF file efficiently so it will be faster. |
|
310 ; Try to predefine everything before you write to it. |
|
311 |
|
312 f = addfile(cdf_file,"c") |
|
313 setfileoption(f,"DefineMode",True) ; Enter predefine phase. |
|
314 |
|
315 ; Write global attributes to file. It's okay to do this before |
|
316 ; predefining the file's variables. We are still in "define" mode. |
|
317 |
|
318 fAtt = True |
|
319 fAtt@description = "Data read in from " + filename + " ASCII file." |
|
320 fAtt@creation_date = systemfunc ("date") |
|
321 fileattdef( f, fAtt ) |
|
322 |
|
323 ; Write dimension names to file. If there are no character variables, |
|
324 ; then there's only one dimension name ("nvalues"). |
|
325 |
|
326 nyear = -1 |
|
327 nmonth = 12 |
|
328 |
|
329 dim_names = (/ "year", "month" /) |
|
330 dim_sizes = (/ nyear , nmonth /) |
|
331 dimUnlim = (/ True , False /) |
|
332 |
|
333 filedimdef( f, dim_names, dim_sizes, dimUnlim ) |
|
334 filedimdef( f, "lat", 1, False ) |
|
335 filedimdef( f, "lon", 1, False ) |
|
336 |
|
337 ; Define each variable on the file. |
|
338 |
|
339 filevardef( f, "year", "integer", "year" ) |
|
340 filevardef( f, "lat" , "float" , "lat" ) |
|
341 filevardef( f, "lon" , "float" , "lon" ) |
|
342 |
|
343 ; Don't deal with variable Month (i=0). |
|
344 |
|
345 do i=0,nfields-1 |
|
346 filevardef(f, var_names(i), var_types(i), dim_names) |
|
347 end do |
|
348 |
|
349 ; Loop through each field, read the values for that field, print |
|
350 ; information about the variable, and then write it to the netCDF |
|
351 ; file. |
|
352 |
|
353 do i=0,nfields-1 |
|
354 ifield = i+1 ; Fields start at #1, not #0. |
|
355 |
|
356 ; Note: you can't write strings to a netCDF file, so these have |
|
357 ; to be written out as character arrays. |
|
358 |
|
359 tmp_data = new((/1,nmonth/),var_types(i)) |
|
360 |
|
361 tmp_data(0,:) = read_field(data,ifield,dindices,var_types(i)) |
|
362 |
|
363 ; tmp_data(0,:) = out_data(:) |
|
364 |
|
365 ; Print some info about the variable. |
|
366 |
|
367 print("") |
|
368 print("Writing variable '" + var_names(i) + "' (field #" + ifield + ").") |
|
369 print("Type is " + var_types(i) + ".") |
|
370 print("min/max = " + min(tmp_data) + "/" + max(tmp_data)) |
|
371 |
|
372 if(any(ismissing(tmp_data))) then |
|
373 print("This variable does contain missing values.") |
|
374 else |
|
375 print("This variable doesn't contain missing values.") |
|
376 end if |
|
377 |
|
378 f->$var_names(i)$ = tmp_data ; Write to netCDF file. |
|
379 |
|
380 delete(tmp_data) ; Delete for next round. |
|
381 ; delete(out_data) |
|
382 end do |
|
383 |
|
384 f->year = year |
|
385 f->lat = lat |
|
386 f->lon = lon |
|
387 end |