How do you vertically concatenate two tables with different variables of different data types

I have two tables with datetime, double, and string variables. I would like to vertically concatenate them.

The tables do not have the same variables, but the ones that do match need to line up in the same columns or it is hard to compare the value in table1 with the value in table2.

I found this thread, but it creates doubles for all of the new variables, which throws an errror with the strings and datetimes.

I tried creating a new table with the same number of rows as table1, and same variables missing from table2. But this requires typing the variables into my script

adder_table = table('Size',sz,'VariableTypes',varTypes,'VariableNames',varNames);

but I don't know how to get the VariableTypes of the difference between the two tables.

I would prefer to not type the names of the variables into my script, as my data sets might change as the initiative adds or drops variables.

Table1

string1	double1	double2
ford	50	0.002439379

Table2

string1	string2	double1	double2	double3
mazda	Sub	30	0.227799138	0.210079407
Tesla	Super	10	0.453158897	0.225137845
ford	none	-10	0.678518656	0.002884727
mazda	Sub	-30	0.903878415	0.290356264

Table_want

string1	string2	double1	double2	double3
ford		50	0.002439379	
mazda	Sub	30	0.227799138	0.210079407
Tesla	Super	10	0.453158897	0.225137845
ford	none	-10	0.678518656	0.002884727
mazda	Sub	-30	0.903878415	0.290356264

[Bora Eryilmaz]

"...but the ones that do match need to line up in the same columns". In your example, which table columns/variables are supposed to match? Given your output (merged) table, I don't see the logic behind your desired match.

[Bora Eryilmaz]

There is a concept of table join. You can learn more about it here: https://www.mathworks.com/help/matlab/ref/table.join.html.

[Ted HARDWICKE]

sorry for the bad example. I revised the original.

[Bora Eryilmaz]

Yup. That is what's called an outer table join: https://www.mathworks.com/help/matlab/ref/table.outerjoin.html.

Answer 1

[Bora Eryilmaz]

C1 = {"ford" 50 0.002439379};
T1 = cell2table(C1, "VariableNames", ["string1" "double1" "double2"]);

C2 = {...
"mazda" "Sub" 30 0.227799138 0.210079407; ...
"Tesla" "Super" 10 0.453158897 0.225137845; ...
"ford" "none" -10 0.678518656 0.002884727; ...
"mazda" "Sub" -30 0.903878415 0.290356264};
T2 = cell2table(C2, "VariableNames", ["string1" "string2" "double1" "double2" "double3"]);

T = outerjoin(T1, T2, 'MergeKeys',true)

T = 5×5 table
    string1    double1     double2      string2      double3 
    _______    _______    _________    _________    _________

    "Tesla"       10        0.45316    "Super"        0.22514
    "ford"       -10        0.67852    "none"       0.0028847
    "ford"        50      0.0024394    <missing>          NaN
    "mazda"      -30        0.90388    "Sub"          0.29036
    "mazda"       30         0.2278    "Sub"          0.21008

[Peter Perkins]

Bora is correct that outerjoin can be used to do this specific example. However, if the two input tables "overlap", i.e. the same combination of "key" values appears in a row in both tables, things can start getting funny, depending on what you want to happen. If you know that's not the case, you are OK.

[Ted HARDWICKE]

Thanks for the recommendation. The outerjoin works. I should have recalled that option.

Peter is correct. The results with the real tables which are much larger yeild different results depending on the order. But understanding of that is available in the documentation and trial and error.

[Stephen23]

@Ted HARDWICKE: please remember to accept this answer if it helped you!