Perform conditional operations on column of doubles

My preferred method to solving this problem would be an if or for argument measuring the length doubles within a column and performing either *10^-5 or *10^-4 depending on the result.

What I've done so far in this code is read a timestamp that is a serial date, converted it into a string, split the string at the decimal, split the date portion and time portion into separate columns. Now, I need to convert those numbers (doubles) into comparable formats for date and time that I have in another array. The format of that array is Date= 7, 8, 9 or 10 and time= 0.69123 (traditional day number for date and serial time for time). This is what I have so far:

    data_compiled= compiled_data_import('data compiled.csv');
    timestamps= cellfun(@(x) num2str(x,15),data_compiled(2:76944,1),'UniformOutput', false);
    dateandtime= cellfun(@(x) strsplit(x,'.'),timestamps,'UniformOutput', false);
    date_time= str2double(vertcat(dateandtime{:}));
    date= date_time(:,1)-42185;
    time= date_time(:,2)*10^-5;

the intermittent outputs look something like this:

timestamps= 42192.6192700

dateandtime= <42192,6192700>

date_time= 42192 6192700

date= 7

time= 0.61927

The problem arises when the time portion of the time stamp has precision<5. This causes the string to be shorter when I use num2str, and therefore 1 order off when I use str2double. My preferred method to solving this problem would be an if or for argument measuring the length of the double and performing either *10^-5 or *10^-4

Answer 1

[Walter Roberson]

You can specify a format string for num2str instead of a number of digits. For example,

num2str(x, '%.15f')