"Error using textscan: Badly formed format string" with R2012b

Hello,

I want to read about 14 million rows of data from an .csv file an save it in a cell array.

The .csv file looks like this:

vehicle_id;_date;measurement_start;measurement_end
100454;2016-09-11T15:01:14.769Z;2.620;6.400
100454;2016-09-11T15:01:15.769Z;7.820;10.400
100452;2016-09-11T15:01:16.769Z;3.620;9.400
100453;2016-09-11T15:01:17.769Z;4.620;5.400
100454;2016-09-11T15:01:18.769Z;20.620;25.400
100458;2016-09-11T15:01:19.769Z;8.620;16.400
100454;2016-09-11T15:01:20.769Z;45.620;50.400

This is my Script to create the cell Array:

clear;
[filename, pathname] = uigetfile({'*.csv','All Files (*.*)'}, 'Pick a file');
filename = [pathname filename];
delimiter = ';';
startRow = 2;
formatSpec = '%f %{yyyy-mm-ddTHH:MM:SSZ}D %f %f %*{^\n}';
fileID = fopen(filename,'r');
dataArray = textscan(fileID, formatSpec, 'Delimiter', delimiter, 'HeaderLines' ,startRow-1, 'ReturnOnError', false);
fclose(fileID);

The mistake is in formatSpec when I try to format the Date and Time.

It's one easy mistake, but I can't figure it out.

Answer 1

[Jan]

Use this instead:

formatSpec = '%f %s %f %f %*{^\n}';

and convert the dates afterwards using datenum. I do not remember if datenum of R2012b can handle this format, but in case of troubles or if this is time critical, you can use FEX: DateStr2Num (use the format spec 31).

Answer 2

[Walter Roberson]

The %D format spec was not supported until R2013b.

[Simon Mürwald]

This explains a lot, I looked at every function if it is supported, but it didn't came to my mind that %D was not supported :/

If I replace the %D with %s everything works fine, but I get a cell array that looks like this:

dataArray = 

      [7x1 double]    {7x1 cell}    [7x1 double]    [7x1 double]

Is there a way to get it look like this:

dataArray = 

    [100454]    '2016-09-11T15:01:14.769Z'    [ 2.6200]    [ 6.4000]
    [100454]    '2016-09-11T15:01:15.769Z'    [ 7.8200]    [10.4000]
    [100452]    '2016-09-11T15:01:16.769Z'    [ 3.6200]    [ 9.4000]
    [100453]    '2016-09-11T15:01:17.769Z'    [ 4.6200]    [ 5.4000]
    [100454]    '2016-09-11T15:01:18.769Z'    [20.6200]    [25.4000]
    [100458]    '2016-09-11T15:01:19.769Z'    [ 8.6200]    [16.4000]
    [100454]    '2016-09-11T15:01:20.769Z'    [45.6200]    [50.4000]

then I could sort it without transforming it.

[Walter Roberson]

[num2cell(dataArray{1}), dataArray{2}, num2cell(dataArray{3}), num2cell(dataArray{4})]