Speeding up data processing for large data set

Hello,

Would anybody be able to help me to speed up this code? It simply processes data by looking for a change in table values (initTab.Val) from a 1 to a 0 then places the corresponding time value (init.Time) in a different table.

The datatypes for both values are double.

The table size could be upto 40e6 * 2

The table I'm currently working on is 6e6 *2

I had a little experiment to find if the time the code takes to execute varies linearly with the initial table size. a table height of 50000 took 7 seconds whereas it took nearly 2 minutes for 500000, so not quite linear.

My computer has plenty of CPU and RAM leftover, whilst running, only 30% of each is being used.

StartLine=20;

initTabHeight=height(initTab);

Tab2Height=initTabHeight/2;%find table length

Tab2SZ=[Tab2Height,2];%table dimensions:height by width

Tab2VarTypes=["double","double"];

Tab2=table('Size',Tab2SZ,'VariableTypes',Tab2VarTypes);%create table

PreVal=initTab.Val(StartLine);%defining 1st value to be compared in below for statement

Tab2LineNum=1;

for i= StartLine+1:initTabHeight %cuts off last line

CurrentVal=initTab.Val(i);

if PreVal > CurrentVal

Tab2{Tab2LineNum,1}=initTab.Time(i);

Tab2LineNum=Tab2LineNum+1;

end

PreVal=CurrentVal;

i

end

Tab2([Tab2LineNum:Tab2Height],:)=[];

[Mathieu NOE]

hello

for loops are usually a bottleneck

you can chek that with the code profiler

would try to remove the for loop