How to make a line plot and indicating data gap?

Dear,

I have the following datafile (please see the attachement), the data consist of three columnes, Date (YYYY-MM-DD), Time (HH:MM) and the last column is the value which need to be ploted against time. But there is gap in the data for example there is no data between 2020-10-10 to 2020-11-02 and i dont want to connect the line, but instead, i would rather leave a shaded gap and mark it as no data. Do you have any suggestions?

I have tried this code but i had problem combining the data and time; am i on the right path?

data = readtable('data.csv');
dateColumn = data(:, 1);
timeColumn = data(:, 2);
valueColumn = data(:, 3);

% Convert date and time components to serial date numbers and here i get
% error
dateNumbers = datenum(dateColumn{:}, 'yyyy-mm-dd');
timeNumbers = datenum(timeColumn{:}, 'HH:MM');
dateTime = datetime(dateNumbers + timeNumbers, 'ConvertFrom', 'datenum');

plot(dateTime, valueColumn{:});
xlabel('Date and Time');
ylabel('Value');
title('Financial Data');
Example customization options:
grid on;  
datetick('x', 'yyyy-mm-dd HH:MM', 'keeplimits');  
legend('Value'); 

Answer 1

[Sulaymon Eshkabilov]

Here is how it can be done. Note that there are only 4 individual data points are missing and thus, they can be displayed individually.

data = readtable('https://www.mathworks.com/matlabcentral/answers/uploaded_files/1390854/Data.txt');

Warning: Column headers from the file were modified to make them valid MATLAB identifiers before creating variable names for the table. The original column headers are saved in the VariableDescriptions property.
Set 'VariableNamingRule' to 'preserve' to use the original column headers as table variable names.

%data = readtable('Data0.txt');
dateColumn = data(:, 1);
timeColumn = data(:, 2);
valueColumn = data(:, 3);

% Convert date and time components to serial date numbers and here i get
% error
dateNumbers = datenum(dateColumn.date);
timeNumbers = datenum(timeColumn.hr_min, 'HH:MM');
dateTime = datetime(dateNumbers + timeNumbers, 'ConvertFrom', 'datenum');
% Missing data point indices are located
IND = find(ismissing(valueColumn.value));
% Missing data points are taken to be mean of the data in order to display
% in the plot figure
valueColumn.value(IND) = mean(valueColumn.value, 'omitnan');

figure
plot(dateTime(IND), valueColumn.value(IND), 'ro','MarkerFaceColor','y', 'markersize',13);
hold on
plot(dateTime, valueColumn.value, 'b-.');

xlabel('Date and Time');
ylabel('Value');
title('Financial Data');
%Example customization options:
grid on;  
hold off
% datetick('x', 'yyyy-mm-dd HH:MM', 'keeplimits');  
legend( 'Missing data', 'All existing data'); 

[Harr]

Dear Sulaymon,

Thank you very much for your effort, unfortunately i cannot run it (is it because i use Matlab 2018b?) and something wrong with the year as well "4043-4044" ! Do you know what was wrong?

/Harr

[Sulaymon Eshkabilov]

Use these commands:

data = readtable('https://www.mathworks.com/matlabcentral/answers/uploaded_files/1390854/Data.txt');

Warning: Column headers from the file were modified to make them valid MATLAB identifiers before creating variable names for the table. The original column headers are saved in the VariableDescriptions property.
Set 'VariableNamingRule' to 'preserve' to use the original column headers as table variable names.

dateColumn = data(:, 1);
timeColumn = data(:, 2);
valueColumn = data(:, 3);
dateNumbers = datenum(dateColumn.date);
timeNumbers = datenum(timeColumn.hr_min, 'HH:MM');
dateTime = datetime(dateNumbers, 'ConvertFrom', 'datenum');

% Missing data point indices are located
IND = find(ismissing(valueColumn.value));
% Missing data points are taken to be mean of the data in order to display
% in the plot figure
valueColumn.value(IND) = mean(valueColumn.value, 'omitnan');

figure
plot(dateTime(IND), valueColumn.value(IND), 'ro','MarkerFaceColor','y', 'markersize',13);
hold on
plot(dateTime, valueColumn.value, 'b-.');

xlabel('Date and Time');
ylabel('Value');
title('Financial Data');
%Example customization options:
grid on;  
hold off
% datetick('x', 'yyyy-mm-dd HH:MM', 'keeplimits');  
legend( 'Missing data', 'All existing data'); 

Answer 2

[Seth Furman]

datestr is discouraged

I should mention that datestr is discouraged. Prefer datetime where possible.

For example,

dt = datetime("2020-10-01","Format","uuuu-MM-dd")

dt = datetime
   2020-10-01

dt.Year, dt.Month, dt.Day

ans = 2020

ans = 10

ans = 1

Consider xregion in MATLAB R2023a or later

% Read data as a timetable
tt = readtimetable("Data.txt", TextType="string", VariableNamingRule="preserve")

tt = 20844×2 timetable
       date       hr:min     value 
    __________    _______    ______

    2020-10-01    "01:00"    9.6179
    2020-10-01    "01:26"    9.6321
    2020-10-01    "01:30"      9.73
    2020-10-01    "01:46"      9.79
    2020-10-01    "02:00"    9.8258
    2020-10-01    "02:30"    9.8682
    2020-10-01    "03:00"    9.7716
    2020-10-01    "03:23"    9.7292
    2020-10-01    "03:30"    9.4418
    2020-10-01    "03:43"    8.8921
    2020-10-01    "03:50"    9.1016
    2020-10-01    "04:00"    9.1988
    2020-10-01    "04:12"    9.2653
    2020-10-01    "04:30"    9.3416
    2020-10-01    "05:00"    9.2477
    2020-10-01    "05:13"     9.203

% Convert duration strings to duration objects
tt.("hr:min") = duration(tt.("hr:min"), InputFormat="hh:mm")

tt = 20844×2 timetable
       date        hr:min     value 
    __________    ________    ______

    2020-10-01    01:00:00    9.6179
    2020-10-01    01:26:00    9.6321
    2020-10-01    01:30:00      9.73
    2020-10-01    01:46:00      9.79
    2020-10-01    02:00:00    9.8258
    2020-10-01    02:30:00    9.8682
    2020-10-01    03:00:00    9.7716
    2020-10-01    03:23:00    9.7292
    2020-10-01    03:30:00    9.4418
    2020-10-01    03:43:00    8.8921
    2020-10-01    03:50:00    9.1016
    2020-10-01    04:00:00    9.1988
    2020-10-01    04:12:00    9.2653
    2020-10-01    04:30:00    9.3416
    2020-10-01    05:00:00    9.2477
    2020-10-01    05:13:00     9.203

% Move duration data to timetable row-times
tt.date = tt.date + tt.("hr:min");
tt.date.Format = "uuuu-MM-dd HH:mm"; % Update datetime format to include hours and minutes
tt.("hr:min") = [] % Remove "hr:min" variable, which is no longer needed

tt = 20844×1 timetable
          date          value 
    ________________    ______

    2020-10-01 01:00    9.6179
    2020-10-01 01:26    9.6321
    2020-10-01 01:30      9.73
    2020-10-01 01:46      9.79
    2020-10-01 02:00    9.8258
    2020-10-01 02:30    9.8682
    2020-10-01 03:00    9.7716
    2020-10-01 03:23    9.7292
    2020-10-01 03:30    9.4418
    2020-10-01 03:43    8.8921
    2020-10-01 03:50    9.1016
    2020-10-01 04:00    9.1988
    2020-10-01 04:12    9.2653
    2020-10-01 04:30    9.3416
    2020-10-01 05:00    9.2477
    2020-10-01 05:13     9.203

% Plot timetable data and add an xregion with legend entry labeling it as
% having no data
plot(tt, "value");
xregion(datetime(2020,10,10), datetime(2020,11,02), DisplayName="No data");
xtickformat("uuuu-MM-dd HH:mm");
xtickangle(45);
legend(["value" "No data"]);