Import Data and Analyze with MATLAB 

From where we can get a data file

@Bushra Noor-ul-ain i am getting error can you help

🤣Glad to hear that it helped!

Thanks!

I'm glad it worked for you.

Thanks!

+Zied Nouir Mahjoub I don't have a video on SQL statements but here is some information online: www.mathworks.com/help/database/run-sql-query.html and www.mathworks.com/help/database/ug/create-a-query-using-a-matlab-variable.html

I'd recommend the reshape function: www.mathworks.com/help/matlab/ref/reshape.html

+ray ho, look at functions such as strfind and strcmp. Once you find the name, you can look up the associated value or use the index to go onto the next step.

+Saeed Muhammad, you have a couple different options. (1) create an interpolation of your data and compare the functions that fit your data (see apmonitor.com/che263/index.php/Main/MatlabDataRegression or www.mathworks.com/help/matlab/ref/spline.html). (2) compare nearest data points by finding one value that is close to the time in the other list (see stackoverflow.com/questions/8089031/how-do-i-find-values-close-to-a-given-value)

thanks allot sir for yours reply

+kisore hari It is probably because the file import wasn't successful. Take a look at the companion video to this one (on Python), starting here: ru-vid.com/video/%D0%B2%D0%B8%D0%B4%D0%B5%D0%BE-Tq6rCWPdXoQ.htmlm38s This is a MATLAB tutorial but you appear to be using Python.

+ABDUL ALIM IBRAHIM I don't have a tutorial on ANNs specifically but here are some on linear and nonlinear regression: apmonitor.com/che263/index.php/Main/MatlabDataRegression

+Fiona Cameron, you can do a linear regression (apmonitor.com/che263/index.php/Main/MatlabDataRegression) and then set y=0 and solve for x in the equation for a line: y=slope x + intercept.

There are multiple courses that teach regression concepts (basic to advanced) here: apm.byu.edu I recommend starting with this material: apmonitor.com/che263/index.php/Main/MatlabDataRegression

I recommend that you try this function for autocorrelation: www.mathworks.com/help/econ/autocorr.html I also like to use the svd function to determine the degree of correlation between the signals. The relative magnitude of the singular values is the thing that you want to use to determine if the signals are unique versus just scaled values of each other. Here is help on the fft function: www.mathworks.com/help/matlab/ref/fft.html

Here is information on obtaining a model from data: apmonitor.com/do/index.php/Main/ModelIdentification

hamza albazr One thing that I didn't show in the tutorial is how to create an inline function. Here is one example of plotting derived variables where the dependent variables may not have been known before the function was defined. y = @(t) t.^2; % create a function y(t) = t^2 t = linspace(0,10); % create a vector t with 100 points between 0 and 10 plot(t,y(t)); % create a plot with t and y(t) In the case of the tutorial video, you could select a sub-set of the times (first 5 rows) with time = sensors(1:5,1), s1 = sensors(1:5,2), s2 = sensors(1:5,3), s3 = sensors(1:5,4) and then an average with avg = @(v1,v2,v3) (v1+v2,v3)/3. You should be able to plot with the command plot(time,avg(s1,s2,s3)). Please let me know if this doesn't answer your question or if you have another question.

Try something like this: Your_array{1:10:end,2}

+BlinkTF2, for functions you should use the addpath command. Place your functions in a directory and then include something like addpath('c:/my_functions'). You should then be able to use the functions in that directory.

Islam, you may need to transpose one of your data vectors so that you have the same number of rows. You can see the size of each thing you are trying to plot with the "size" function: www.mathworks.com/help/matlab/ref/size.html. Make sure that x and y have the same dimensions. You can use the command "hold on" to plot multiple data sets on the same plot.

Try adding an ‘ to the vector. It’s probably because one is a column and one is a vector

Rasika Dilhani You probably need to select in Excel the Save As option and save the file as a txt or csv file. There is also a xlsread function in MATLAB if you'd like to read an Excel file directly: www.mathworks.com/help/matlab/ref/xlsread.html

APMonitor.com Thank you so much. Now it works properly. I waste too much my time to do this. Your guideline help me to success this matter.

You can read a data column with disp(x(:,2)) for the second column. The : indicates the second column.

Mehfooz Ali, here is some help on importing data into Simulink: apmonitor.com/che436/index.php/Main/ProjectProgramming (see the first example).

What type of TTL converter are you trying to use? Here is something I found online: "UARTs transmit one bit at a time at a specified data rate (i.e. 9600bps, 115200bps, etc.). This method of serial communication is sometimes referred to as TTL serial (transistor-transistor logic). Serial communication at a TTL level will always remain between the limits of 0V and Vcc, which is often 5V or 3.3V."

Check out the optimization course material at apmonitor.com/me575 under the application project section. There was one project that was similar to yours at apmonitor.com/me575/uploads/Main/2013_Knee_Simulator.pdf

This is easy with Python or NodeRed but I haven't tried it with MATLAB. Here is another question that you may want to follow to see if Mathworks provides an answer: www.mathworks.com/matlabcentral/answers/130100-sentiment-analysis-and-opinion-mining

+Youssef Zakaria, one way to avoid this is to import the data as a cell array and then you don't need to differentiate between numbers and text. www.mathworks.com/help/matlab/cell-arrays.html

You could create a loop and select only those rows that have that condition. Alternatively, it is easy to do this with Python (Pandas package): apmonitor.com/pds/index.php/Main/CleanseData

Alaan, check out the ANN (artificial neural network) toolbox. It will help you fit a model. However, you may find that an ANN model extrapolates poorly. With such a small system (2 x 2) you may find other forms that extrapolate better such as a first principles model or better empirical approaches. Make sure you keep a separate data set for validation.

+ABDUL ALIM IBRAHIM How many inputs and outputs do you have in your data? If you only have 1 or 2 outputs (and 10 or 9 inputs) then it may be possible to model. With only 45 data points, the results may not be very good because there can be more unknown parameters than data points. Is there a possibility to get more data?

+APMonitor.com Thank you for your responce sir, it is possible for me to attached the data for you to see it yourself? its a data collected from ABU, Zaria, teaching Hospital for Short Insulin Tolerance Test, I want to identify the model from the data to analyse Glucose-insulin kinetics. Thank you

+ABDUL ALIM IBRAHIM Your data looks very interesting. You may be able to do an analysis similar to the one found here: www.ncbi.nlm.nih.gov/pmc/articles/PMC2628038/ We also have some mechanistic models for Type-I diabetes but I'm assuming that the data you sent were for Type-II diabetics. What is the dosage of insulin delivered as part of the test? You need this information as well to develop a model from the step test data.

+Shikha Singh, here are some functions that may help you with imaginary numbers: www.mathworks.com/help/matlab/complex-numbers.html I'd recommend that you enter debug mode and step through the first couple cycles of your loop to have a more detailed look at what is changing.

there is an error in this statement E = [elec(:,4)+elec(:,5)*i elec(:,6)+elec(:,7)*i elec(:,8)+elec(:,9)*i]; instead of giving ans in terms of a+bi, its is giving only a how to overcome this problem?

Shikha Singh, is this statement in a loop? Your syntax looks correct to me. One problem could be if "i" is defined as your loop variable then you are redefining the definition of an imaginary number. Either use a different iterator in your "for" or "while" loop or else switch to "j" for your imaginary number.

I have to store 80 values in an array with 10 rows and 8 columns using statement poyndb = 10*log10(modu) - 3 array_poyndb(k) = poyndb; and then finding average using array_poyndb(k) = poyndb; could you tell me is there a problem in this statement?

you can transpose the data after importing data = data'

Thank you Sir. If I have further question, I can still ask.

Try copying the code from here: apmonitor.com/che263/index.php/Main/MatlabPlots specifically at apmonitor.com/che263/index.php/Main/MatlabPlots?action=sourceblock&num=1 or apmonitor.com/che263/index.php/Main/MatlabPlots?action=sourceblock&num=2

You can get the source files for this problem from: apmonitor.com/che263/uploads/Main/matlab_data_analysis.zip

+Asif Nizamani, Here is an example Kalman filter script: www.mathworks.com/matlabcentral/fileexchange/5377-learning-the-kalman-filter

Maybe a Google search?

Je ne trouve pas la base de données de roulement

@@mezourachahinez5721 Does this help? www.ntnamericas.com/en/website/documents/brochures-and-literature/catalogs/ntn_a1000xi_ball_and_roller_bearings_lowres.pdf

I'd recommend loading each file individually and then perform the training and testing / validation. Alternatively, you can load one file and then parse out the training and test data. Here is a related example but in Python: apmonitor.com/do/index.php/Main/DeepLearning

There are many good resources online such as www.mathworks.com/help/matlab/matlab_external/choose-c-or-c-compilers.html

After you import the data files, you can combine them with combined_data = [data1 data2 data3 data4 data5] for column oriented data or combined_data = [data1; data2; data3; data4; data5] for row oriented data.

Try a meshgrid www.mathworks.com/help/matlab/ref/meshgrid.html I'm not sure about images

Try imopen: www.mathworks.com/help/images/ref/imopen.html

This may help: www.mathworks.com/matlabcentral/answers/29099-importing-formated-time-data-into-matlab

See the 3D Numpy arrays for information on how to do this with Python: apmonitor.com/dde/index.php/Main/PythonNumpy

You may have headers that interfere with the numeric value read. I'd recommend either deleting the header row or else looking at a forum post where it discusses reading a text file with headers.

Does this help? www.mathworks.com/matlabcentral/answers/27734-how-can-i-skip-lines-when-loading-a-text-file

There is an example here: apmonitor.com/heat.htm

+Ayush Sharma, you may need to use a different method such as textscan to read in the date information: www.mathworks.com/help/matlab/import_export/import-formatted-dates-and-times.html Once you import the data, you can use one of these methods to lookup the row: www.mathworks.com/matlabcentral/answers/218033-efficient-method-for-finding-index-of-closest-value-in-very-large-array-for-a-very-large-amount-of-e

+soumendra singh One way to do this is to indicate the missing data as NaN values. Then you can apply the nanmean function to average over the data, ignoring the NaN data (www.mathworks.com/help/stats/nanmean.html). If you don't know if a particular number is missing or not and want to do this as a loop, then use the isnan function to determine which elements are numbers (www.mathworks.com/help/matlab/ref/isnan.html)

Sorry, I don't have experience with COE files.

Here is an example to help get you started: apmonitor.com/do/index.php/Main/DeepLearning

Oh soo much Thank you!!!

+Tareq Atiany There is an example of noise, outliers, and drift for contamination of data on the dynamic optimization course web-site here: apmonitor.com/do/index.php/Main/DynamicData (see the example at the bottom).

I recommend that you take a look at the ANN toolbox in MATLAB and the commands related to graycomatrix.

okay thanks sir

Here is a link that helps with Excel: www.mathworks.com/discovery/matlab-excel.html

+Chinomso Uche do you also have equations or just data? You can either use a data driven empirical approach where the equation form is determined at the time you perform the regression (e.g. artificial neural network), a correlation that is based on first principles (e.g. ab initio calculations), or a combination of the two.

+APMonitor.com I want to run a simple model to simulate the permeability measurement using a given parameters n data. Some of the parameters are constant since I ran the experiment on one core sample but for five time where I had to Chang the flow rate which ended up having a change in the value of the change in pressure and the permeability. So which is the best way to model this on Matlab, am to compare the real values gotten from the experiment and the simulated values from the Matlab.

Here is information on Regression with MATLAB: apmonitor.com/che263/index.php/Main/MatlabDataRegression

Bashir Muhammad Take a look at the SVM documentation here: www.mathworks.com/help/stats/svmtrain.html It appears that there is a particular input format that is required.

+nor amira abu bakar I'm not familiar with your code. Maybe a paper like ieeexplore.ieee.org/xpl/login.jsp?tp=&arnumber=6559645 can help you.

I'm not familiar with the SPM package - is it for analysis of brain images? It appears that the versions available here www.mathworks.com/matlabcentral/linkexchange/links/977 have newer versions (SPM8 or SPM12) - maybe try the newer versions.

APMonitor.com thank you for attention and for helpful videos ☺️

A common error is that the text file has a header and is not just numbers. You may want to try textread (www.mathworks.com/help/matlab/ref/textread.html) or else remove the headers.

I removed the header just kept the numbers in the text file, but its showing like this ??? Error using ==> load Number of columns on line 1 of ASCII file C:\Users\SHIKHA\Desktop\poynting\Monopole800_FFefe.txt must be the same as previous lines.

That is a strange error because line 1 doesn't have any previous lines. I'd recommend that you open the file with a program such as Notepad++ and turn on the option to view all characters. Make sure your numbers are separated by commas or tabs as well.

That error has gone. Could you please tell me to how Plot module and phase in polar coordinates of poynting vector? Which commands to use?

Shikha Singh, check out the programming class at apmonitor.com/che263 There is a section on generating plots in Matlab.

Sorry, I don't have datasets for Wireless Sensor Networks.

Try the reshape function: www.mathworks.com/help/matlab/ref/reshape.html

+Neha Gupta, here is the top Google search for your request: sccn.ucsd.edu/~arno/fam2data/publicly_available_EEG_data.html

It was not worked when tried to import It have not showed any xlsx file only mat files shown

apmonitor.com/che263/index.php/Main/MatlabDataAnalysis

Sorry, I don't know about that one.

Here is information on regression: apmonitor.com/che263/index.php/Main/MatlabDataRegression

Here are instructions for building your own EEG: www.instructables.com/id/DIY-EEG-and-ECG-Circuit/ or here are some publicly available datasets: sccn.ucsd.edu/~arno/fam2data/publicly_available_EEG_data.html The data appears to be in an rdf format and you'll need a program such as sccn.ucsd.edu/wiki/A01:_Importing_Continuous_and_Epoched_Data to import it. If you can convert it to a text file then you can use MATLAB to manipulate the data or maybe use a program such as EEGLAB for MATLAB: sccn.ucsd.edu/eeglab/

Does it work to upload your Excel file to drive.matlab.com so that you can import it?

You can type Ctrl-h to open up the dialog box for Find and Replace.

@@apm Thank you. So I have to change each value separately?

@@muqtasidaliawan no - you can select "Replace All"

@@apm Thank you so much for your reply. It will help. 👌 👍

+toukal manel A text file is probably the easiest method to import data. You can also access the database directly: www.mathworks.com/products/database

You can call a Python function from Matlab. Try this package: pypi.org/project/traceanalyzer/

+Ibrahm Musa Here is a link on LDA: www.mathworks.com/matlabcentral/fileexchange/29673-lda--linear-discriminant-analysis and PCA: www.mathworks.com/help/stats/pca.html

One other resource that may help you as a beginner is to visit the Matlab section of my course on computational tools apmonitor.com/che263 Please navigate to the Matlab section for introductory videos.

+Dzulfadhly Shaari, try: mean (data(:,2)) to get the average of column 2.