Hi, this may seem a little off the topic but does the batch language allow for commands like strcat, strcat, printf etc. For instance I want to parition data according to a consensus amino acid sequence. So I have the consensus amino acid sequence in a {site x amino acid} matrix as follows:

{ { 0, 1, 2, 3, 4}, { 14, 2, 5, 14, 5} }

which i can sort by column 1 to:

{ { 1, 2, 4, 0, 3 }, { 2, 5, 5, 14, 14 } }

so using a for and while loop i can identify which sites all have the same consensus amino acid. ie. site 1, sites 2 & 4, sites 0 & 3.

Now I want to create filters for each consensus amino acid (each of which will get a separate model). So some way of doing:

DataSetFilter myFilter1 = CreateFilter ( myData, 1, "1" ); /* AA 2 */
DataSetFilter myFilter2 = CreateFilter ( myData, 1, "2,4" ); /* AA 5 */
DataSetFilter myFilter3 = CreateFilter ( myData, 1, "0,3" ); /* AA 14 */

So I thought a for loop that will concatenate values in a matrix as a string, where the matrix is only retained for each AA. ie for AA 5
matrix = { 2, 4 }

stringVar = "";
for ( i = 0; i < k; i = i + 1 ) { /* k is the number of sites with AA 5 */
    strcat ( stringVar, matrix [ i ] );  
    if ( i != k-1 ) {
        strcat ( stringVar, ", " );
    }
}

and then theoretically i can call this string with the CreateFilter?

DataSetFilter myFilter2 = CreateFilter ( myData, 1, stringVar ); /* AA 5 */

and finally the other problem would be how to index the myFilter in a forward loop...myFilter1, myFilter2, .... for 20 amino acids. i am pretty new to both HyPhy and the batch language so please let me know if i'm completely off the wall with this ?. thanks ./w :o

For strings, the '+' operation will concatenate the strings. A more efficient way to do this is to create a "buffer" string like this

Code (]

stringBuffer = ""; /* declare as string */
stringBuffer * 128; /* mark as buffer string, i.e. it will autogrow as stuff is pushed on it */
stringBuffer * ("some text");
stringBuffer * ("some more text");
...
stringBuffer * 0; /* complete the buffering; free unused memory and such */

[/code):

Still, for what you do, a more efficient way to create the filter is to use something like this. Suppose 'indexMat' holds the amino-acid value in the consensus at the that position (row 2 in your example). Then call something like [code] indexMat =[ 0, 1, 2, 3, 4}, { 14, 2, 5, 14, 5} }; DataSetFilter myFilter3 = CreateFilter ( myData, 1, indexMat[1][siteIndex]==14 )

What this will do is evaluate the expression indexMat[1][siteIndex]==14 for every siteIndex from 0 to myData.sites-1, and put all sites for which the expression is true (>0) into the filter.

HTH,
Sergei

thanks Sergei, your reply is once again informative and helpful. your efforts are really appreciated. cheers ./w