How to replace char 'null' in number 0?

Hi! I`ve got a problem with using plot function cause there is some 'null's (char type) in my data files . So i would like to change/replace that nulls in 0 (num) . Could someone help to find useful function or to offer other solution.

[Bizzy Dy]

how it set in tsv data file

[Guillaume]

It would be much simpler to fix the way you're reading the file rather than trying to fix the problem after the fact.

For that we would need to see the code you're using (not a screenshot, an actual copy as text) and a sample file (again not a screenshot!)

[Bizzy Dy]

%%read
clear all, close all;
h=tdfread('result_3259_m.tsv','tab')

%%get
time=getfield(h,'time')
Qus=getfield(h,'input0x2Eflowrate0x2Eflowmeter')
P1=getfield(h,'input0x2Epressure0x2Ebefore')
P2=getfield(h,'input0x2Epressure0x2Eafter')
P3=getfield(h,'input0x2Epressure0x2EforFlowmeter')
T1=getfield(h,'input0x2Etemperature0x2Ebefore')
T2=getfield(h,'input0x2Etemperature0x2Eafter')
A=getfield(h,'input0x2Evalve0x2Eposition')
% bezm=getfield(h,'BEZM')
% rmse=getfield(h,'RMSE')

%%dateformat
formatIn='yyyy-mm-ddTHH:MM:ssZ'
% ds=datetime(time,'InputFormat','yyyy-mm-ddTHH:MM:ssZ')
    dn=datenum(time,formatIn)

%%draw
dateborder1=datenum(2018,07,01,0,0,0)
dateborder2=datenum(2018,07,26,0,0,0)

subplot(2,1,1);
    plot(dn,P1,'y');hold on;
    plot(dn,P2,'c');hold on;
    plot(dn,P3);hold on;
    plot(dn,T1,'m');hold on;
    plot(dn,T2,'g');hold on;
    plot(dn,A,'r');hold on;

datetick('x','dd.mm, HH:MM','keepticks')
datetickzoom
    axis([dateborder1, dateborder2, -20 , 100])
        legend('P1','P2','P3','T1','T2','A')
            grid on
%             grid minor
subplot(2,1,2);
%     plot(dn,bezm);hold on;
    plot(dn,Qus,'r');hold on;
%   plot(dn,rmse,'g');hold on;

datetick('x','dd.mm, HH:MM','keepticks')
datetickzoom('x',1)
% xticks(0:10:100)
     axis([dateborder1, dateborder2, 0 , max(Qus)])
%          legend('Bezm','Qus')
            grid on
%             grid minor

[Stephen23]

Guillaume is right: fix the problem when it occurs, rather than trying to write hack code to "fix" it later. Upload a sample data file by clicking the paperclip button.

[Bizzy Dy]

done

[Guillaume]

Please do not close questions that have been answered.

Answer 1

[Guillaume]

It's much easier to read your file with readtable (which also does not require any toolbox). It is trivial with readtable to convert the 'null' to 0 and read everything else as number.

t = readtable('result_3259_m.tsv', 'FileType', 'text', 'TreatAsEmpty', 'null', 'EmptyValue', 0);
t.time = datetime(t.time, 'InputFormat', 'yyyy-MM-dd''T''HH:mm:ss''Z''');

For further processing you may even want to convert the table to a timetable

I would also recommend that you do not extract the table columns into variables with meaningless names. Just use the table variables directly:

plot(t.time, t.input_flowrate_bezm, 'DatetimeTickFormat', 'dd.MM, HH:mm');

[Bizzy Dy]

Thank you so much!

Answer 2

[Ameer Hamza]

The image shows a struct which char arrays as fields. You can use strrep() to replace the elements of a char array. Although it will still not work with the plot() function because the elements are char arrays, not numeric values. You might need to use str2num() to convert char values t numeric before using plot().