Excluding empty cells ("") from ascending list with Index function - excel

Question

I have the following Excel table:
A B C
1 Boris 4 *
2 Anna 6 *
3 Uli 5 *
4 Inge 4 *
5 Rudi 3 *
6 Ulla 7 *
7
8
9
:
:
99
*In cells C1 to C6 I am using the matrix formula:
={INDEX(A:A;VERGLEICH(KKLEINSTE(B$1:B$99-ZEILE($1:$99)/9^9;ZEILE(A1));B$1:B$99-ZEILE($1:$99)/9^9;0))}
to get the list sorted by names from the smallest to the the highest number according to column B.
The issue is now that my list has 99 rows (as you can see in the table) but not all of them are filled (as you can see in row 7 - 99 in the table).
Therefore, the formular in cell C1 to C6 shows now the value 0 because "" (empty cell) is the smallest value in the list from B1 to B99.
How do I have to change the formular that it considers all values in column B except for the cells that are empty? (Note: If a cell in column B has the value 0 it should be considered. Only when the cell is empty it should be excluded)
Thanks for any help :-)

Answer 1

Just add a condition to check for numericalness:
=INDEX(A:A,MATCH(SMALL(IF(ISNUMBER(B$1:B$99),B$1:B$99-ROW($1:$99)/9^9),ROWS($1:1)),B$1:B$99-ROW($1:$99)/9^9,0))
In German:
=INDEX(A:A;VERGLEICH(KKLEINSTE(WENN(ISTZAHL(B$1:B$99);B$1:B$99-ZEILE($1:$99)/9^9);ZEILEN($1:1));B$1:B$99-ZEILE($1:$99)/9^9;0))
I have replaced ROW (ZEILE) with ROWS (ZEILEN) as it is a more rigorous choice for SMALL's k parameter:
http://excelxor.com/2014/08/25/row-vs-rows-for-consecutive-integer-generation/
Regards