Curve fit identical to input data

Hi,
I am a newbe to Scilab .
I did a curve fit in Octave, based on digitized data, and wanted to reproduce it in Scilab
What I get in Scilab is an exact reproduction of the data, not a fitted curve. It seems that all data points are knots, even when I use the option "not_a_knot" .
I have updated the text file, so that the x-axis is monotone increasing
I have uploaded the input file and the figure I get in octave.
The .sci file is:

charge3=fscanfMat('charge_10C_3.txt');
d=zeros(size(charge3,1)-1);
d=diff(charge3( :,1));
max(d)
min(d)
figure;
p3= splin(charge3( :,1),charge3( :,2), "not_a_knot");
yyk = interp(charge3( :,1),charge3( :,1), charge3( :,2), p3);
plot(charge3( :,1),charge3( :,2),'b',charge3( :,1),yyk,'g');
legends (["U charge","U fit"],[1 2],opt="ul", fontsize=4);
xlabel (["Soc [-]"],"fontsize",4);
ylabel (["U cell [V]"], "fontsize" ,4);
e=charge3( :,2)-yyk
max(e)