Reading a .txt file

I am trying to read a .txt file with different no. of column in each row. Also first few lines contain header information. First few lines for the files are pasted below

Version 8.0 n19 SBUV data for day 006 2011 (2011/01/06) 
 1288 :Number of records
1) year  day  sec-gmt  Lat  Lon  SZA  Total_Ozone  Reflectivity  Aerosol_Index  Quality_residue  Error_Flag
2) ozone (DU) in 13 layers -- pressure level at the bottom of each layer(atm):
 1.000 0.0631 0.0400 0.0251 0.0158 0.0100 0.0063 0.0040 0.00251 0.00158 0.0010 0.00063 0.00040
3) ozone (PPMV) at 15 pressure levels(hPa):
 0.5 0.7 1.0 1.5 2.0 3.0 4.0 5.0 7.0 10.0 15.0 20.0 30.0 40.0 50.0 

 2011   6  5235 -71.26  -25.20 86.12 276.3  0.755  -0.8 0.032  11
  99.76  46.51  41.39 29.819 22.048 15.650 10.123  5.814  2.846  1.286 0.5925 0.2722 0.2207
  1.400  1.773  2.253  2.994  3.711  4.753  5.222  5.389  5.371  5.059  4.439  4.008  3.617  3.184  2.593
 2011   6  5267 -72.87  -28.35 84.42 279.8  0.797  -1.4 0.031  11
 101.67  47.43  41.91 29.989 22.153 15.722 10.123  5.744  2.816  1.271 0.5663 0.2516 0.2030
  1.290  1.669  2.190  2.975  3.675  4.685  5.186  5.388  5.394  5.084  4.460  4.029  3.657  3.239  2.645

I am interested in reading data starting from 2011. I used fscanf and dlmread. fscanf could not work exactly. Also dlmread read few of the columns of the data. But not the full data. Kindly help me to read the data from this txt file. dlmread was used as M=dlmread(filename,'\t',8,0);

Thanks in advance

Answer 1

[dpb]

Of course fscanf can read it if you work on the formatting sufficiently. With an irregular file structure, though, it'll be prudent to use the facilities of textscan to ease the burden somewhat if trying to read the file on the fly.

I'll take a slightly different tack here, though, since the data you're interested in are regularly spaced and just differ by the number of elements/row.

dlmread can only work on numeric-only data, per the doc--

>> help dlmread
dlmread Read ASCII delimited file.
  RESULT = dlmread(FILENAME) reads numeric data from the ASCII
  delimited file FILENAME.  The delimiter is inferred from the formatting
  of the file.
  ...
  All data in the input file must be numeric. dlmread does not operate 
  on files containing nonnumeric data, even if the specified rows and
  columns for the read contain numeric data only.

I pasted your sample data into a file...therein the linewrap shown in the forum disappeared so I presume that's simply a fignewton of the display.

Then, just read the whole file into memory as an array of cell strings and operate from memory to convert...I used textread as it's so simple and useful despite TMW having relegated it to the status or red-haired stepchild--use textscan or whatever instead if you prefer--

After that, just convert the subsequent lines to character strings selecting each group and let str2num do the hard work for you--

>> f=textread('pooj.dat', '%s', 'delimiter', '\n', ...
              'whitespace', '','headerlines',8);
>> hdr=str2num(char(f(1:3:end)))
hdr =
 1.0e+03 *
Columns 1 through 9
  2.0110    0.0060    5.2350   -0.0713   -0.0252    0.0861    0.2763    0.0008   -0.0008
  2.0110    0.0060    5.2670   -0.0729   -0.0284    0.0844    0.2798    0.0008   -0.0014
Columns 10 through 11
  0.0000    0.0110
  0.0000    0.0110
>> DU=str2num(char(f(2:3:end)))
DU =
Columns 1 through 9
 99.7600   46.5100   41.3900   29.8190   22.0480   15.6500   10.1230    5.8140    2.8460
101.6700   47.4300   41.9100   29.9890   22.1530   15.7220   10.1230    5.7440    2.8160
Columns 10 through 13
  1.2860    0.5925    0.2722    0.2207
  1.2710    0.5663    0.2516    0.2030
>> PPM=str2num(char(f(3:3:end)))
PPM =
Columns 1 through 9
  1.4000    1.7730    2.2530    2.9940    3.7110    4.7530    5.2220    5.3890    5.3710
  1.2900    1.6690    2.1900    2.9750    3.6750    4.6850    5.1860    5.3880    5.3940
Columns 10 through 15
  5.0590    4.4390    4.0080    3.6170    3.1840    2.5930
  5.0840    4.4600    4.0290    3.6570    3.2390    2.6450
>>

Done! :)

[Pooja]

Can u pls let me know in

DU=str2num(char(f(2:3:end)))

what is the meaning using 3 here?

[dpb]

Read

doc colon % maybe???

Or, try

(2:3:10)

at the command line and see a numeric example.