Replace NA value with next column value

[YoussefBelloum]

Hello everyone,

hope you're doing well !

please find below my datasctructure for today's problem:

colonne1	colonne2	colonne3	colonne4	colonne5	colonne6	colonne7
NA	ligne1.colonne2	ligne1.colonne3	ligne1.colonne4	ligne1.colonne5	ligne1.colonne6	ligne1.colonne7
ligne2.colonne1	NA	ligne2.colonne3	ligne2.colonne4	ligne2.colonne5	ligne2.colonne6	ligne2.colonne7
ligne3.colonne1	ligne3.colonne2	NA	ligne3.colonne4	ligne3.colonne5	ligne3.colonne6	ligne3.colonne7
ligne4.colonne1	ligne4.colonne2	ligne4.colonne3	NA	ligne4.colonne5	ligne4.colonne6	ligne4.colonne7
ligne5.colonne1	ligne5.colonne2	ligne5.colonne3	ligne5.colonne4	NA	ligne5.colonne6	ligne5.colonne7
ligne6.colonne1	ligne6.colonne2	ligne6.colonne3	ligne6.colonne4	ligne6.colonne5	NA	ligne6.colonne7
ligne7.colonne1	ligne7.colonne2	NA	NA	ligne7.colonne5	ligne7.colonne6	ligne7.colonne7

For every single line, I need to replace every NA value with next column value, like the result below:

colonne1	colonne2	colonne3	colonne4	colonne5	colonne6	colonne7
ligne1.colonne2	ligne1.colonne3	ligne1.colonne4	ligne1.colonne5	ligne1.colonne6	ligne1.colonne7
ligne2.colonne1	ligne2.colonne3	ligne2.colonne4	ligne2.colonne5	ligne2.colonne6	ligne2.colonne7
ligne3.colonne1	ligne3.colonne2	ligne3.colonne4	ligne3.colonne5	ligne3.colonne6	ligne3.colonne7
ligne4.colonne1	ligne4.colonne2	ligne4.colonne3	ligne4.colonne5	ligne4.colonne6	ligne4.colonne7
ligne5.colonne1	ligne5.colonne2	ligne5.colonne3	ligne5.colonne4	ligne5.colonne6	ligne5.colonne7
ligne6.colonne1	ligne6.colonne2	ligne6.colonne3	ligne6.colonne4	ligne6.colonne5	ligne6.colonne7
ligne7.colonne1	ligne7.colonne2	ligne7.colonne5	ligne7.colonne6	ligne7.colonne7

like you noticed on the line 7, there is multiple NA values, the idea is to replace them all with next columns values, and all cases are possible (only one NA, multiple NA side by side, multiple NA separated by columns values, etc...)

One solution consists of using A LOT of nested IF statements (A nightmare to write and maintain) so I'm searching for a slightly more efficient solution..

Thank you for your help and have a nice weekend 🙂

Youssef B

[jwjackso · Accepted solution]

I used multiple preload steps to verify that it worked.

Data:
Load SubField(str3,',',1) as colonne1,
SubField(str3,',',2) as colonne2,
SubField(str3,',',3) as colonne3,
SubField(str3,',',4) as colonne4,
SubField(str3,',',5) as colonne5,
SubField(str3,',',6) as colonne6,
SubField(str3,',',7) as colonne7;
Load *,
str2&Repeat(',',7-FieldCount-1) as str3;
load *,
SubStringCount(str2,',') as FieldCount;
Load *,
replace(str1,'NA,','') as str2;
Load colonne1&','&colonne2&','&colonne3&','&colonne4&','&colonne5&','&colonne6&','&colonne7 as str1;
Load * Inline [
colonne1,colonne2,colonne3,colonne4,colonne5,colonne6,colonne7
NA,ligne1.colonne2,ligne1.colonne3,ligne1.colonne4,ligne1.colonne5,ligne1.colonne6,ligne1.colonne7
ligne2.colonne1,NA,ligne2.colonne3,ligne2.colonne4,ligne2.colonne5,ligne2.colonne6,ligne2.colonne7
ligne3.colonne1,ligne3.colonne2,NA,ligne3.colonne4,ligne3.colonne5,ligne3.colonne6,ligne3.colonne7
ligne4.colonne1,ligne4.colonne2,ligne4.colonne3,NA,ligne4.colonne5,ligne4.colonne6,ligne4.colonne7
ligne5.colonne1,ligne5.colonne2,ligne5.colonne3,ligne5.colonne4,NA,ligne5.colonne6,ligne5.colonne7
ligne6.colonne1,ligne6.colonne2,ligne6.colonne3,ligne6.colonne4,ligne6.colonne5,NA,ligne6.colonne7
ligne7.colonne1,ligne7.colonne2,NA,NA,ligne7.colonne5,ligne7.colonne6,ligne7.colonne7
];

[YoussefBelloum]

@jwjackso 

It's awesome !

Thank you for your time

Youssef B