- GROUPFILTER
- SORT, UNIQUE, then indexing.
How to create a group summary of table such that it takes latest values ?
5 vues (au cours des 30 derniers jours)
Afficher commentaires plus anciens
Hello All,
I am having some problems in creating a summary of table.
I will try to explain my problem in detail.
I have a table named "InputSessionTable" which is shown like this. (Current Input)
% Create table
CustomerName = ["Customer1" ; "Customer1" ; "Customer1" ; "Customer1" ; "Customer2" ; "Customer2" ];
ProjectName = ["Project1" ; "Project1" ; "Project1" ; "Project3" ; "Project2" ; "Project2" ];
VehicleName = ["Vehicle1" ; "Vehicle1" ; "Vehicle1" ; "Vehicle3" ; "Vehicle2" ;"Vehicle2" ];
VehicleType = ["VehicleType1" ; "VehicleType1" ; "VehicleType1" ; "VehicleType3" ; "VehicleType2" ; "VehicleType2"];
EngineerName = ["Engineer1" ; "Engineer1" ; "Engineer3" ; "Engineer3" ; "Engineer2" ; "Engineer2" ];
Date = [datetime(2023,12,27); datetime(2023,12,29); datetime(2023,12,30) ; datetime(2023,12,31); datetime(2023,12,28) ; datetime(2023,12,30)];
VerNo = [1 ; 2 ; 3 ; 1 ; 1 ;2 ];
InputSessionTable = table(CustomerName,ProjectName,VehicleName,VehicleType,EngineerName,Date,VerNo);
InputSessionTable =
6×7 table
CustomerName ProjectName VehicleName VehicleType EngineerName Date VerNo
____________ ___________ ___________ ______________ ____________ _____________ _____
"Customer1" "Project1" "Vehicle1" "VehicleType1" "Engineer1" "27-Dec-2023" 1
"Customer1" "Project1" "Vehicle1" "VehicleType1" "Engineer1" "29-Dec-2023" 2
"Customer1" "Project1" "Vehicle1" "VehicleType1" "Engineer3" "30-Dec-2023" 3
"Customer1" "Project3" "Vehicle3" "VehicleType3" "Engineer3" "31-Dec-2023" 1
"Customer2" "Project2" "Vehicle2" "VehicleType2" "Engineer2" "28-Dec-2023" 1
"Customer2" "Project2" "Vehicle2" "VehicleType2" "Engineer2" "30-Dec-2023" 2
This table is created from the files I am having in the directory.
Here Customer Name, Project Name, Vehicle Name and Vehicle Type are the unique characteristics of the file.
If you observe first 3 rows that Customer Name, Project Name, Vehicle Name and Vehicle Type is same but their Version number is updated to 3.
Like Wise for row 5 and 6 Version number is updated to 2.
Where as row 4 its unique and does not any newer versions.
I am trying to find a way such that it create below table with "OutputSessionTable". (Desired Output)
OutputSessionTable =
3×7 table
CustomerName ProjectName VehicleName VehicleType EngineerName Date VerNo
____________ ___________ ___________ ______________ ____________ _____________ _____
"Customer1" "Project1" "Vehicle1" "VehicleType1" "Engineer3" "30-Dec-2023" 3
"Customer1" "Project3" "Vehicle3" "VehicleType3" "Engineer3" "31-Dec-2023" 1
"Customer2" "Project2" "Vehicle2" "VehicleType2" "Engineer2" "30-Dec-2023" 2
Now if you observe it takes the latest version of each unique project. This eliminates the rows with same unique chaterstics but older version number.
I have tried to use a function called groupsummary. This helps me in creating a unique chatertics of table.
GroupSummaryTable = groupsummary(InputSessionTable,1:4,"max","VerNo")
When using this function I get an putput like this.
GroupSummaryTable =
3×6 table
CustomerName ProjectName VehicleName VehicleType GroupCount max_VerNo
____________ ___________ ___________ ______________ __________ _________
"Customer1" "Project1" "Vehicle1" "VehicleType1" 3 3
"Customer1" "Project3" "Vehicle3" "VehicleType3" 1 1
"Customer2" "Project2" "Vehicle2" "VehicleType2" 2 2
Now here I am stuck that how can I include Engineer Name and Date column with this table in order to get "OutputSessionTable". Those are in string and datetime format.
Let me know your view on this.
Thanks in Advance
1 commentaire
Réponse acceptée
Dyuman Joshi
le 4 Jan 2024
Modifié(e) : Dyuman Joshi
le 4 Jan 2024
It is helpful to provide/attach data (or a sample of it) with the question, so it is easier for us to work with.
Also, In general, it is better to use the datetime data type to represent any data with dates (or time or a mix of both). However, if you want to keep it as string, you can remove that line.
%Data for table
CustomerName = "Customer" + [1 1 1 1 2 2].';
ProjectName = "Project" + [1 1 1 3 2 2].';
VehicleName = "Vehicle" + [1 1 1 3 2 2].';
VehicleType = "VehicleType" + [1 1 1 3 2 2].';
EngineerName = "Engineer" + [1 1 3 3 2 2].';
Date = ["27-Dec-2023" "29-Dec-2023" "30-Dec-2023" "31-Dec-2023" "28-Dec-2023" "30-Dec-2023"].';
VerNo = [1 2 3 1 1 2].';
T = table(CustomerName,ProjectName,VehicleName,VehicleType,EngineerName,Date,VerNo);
%Convert the Date variable to datetime() data type
%convertvars() can also be used for the same
T.Date = datetime(T.Date, 'InputFormat', 'dd-MMM-yyyy');
%% Sort the rows of the table first by VerNo then by Date, in descending order
T = sortrows(T, {'VerNo', 'Date'}, 'descend')
%% Get the indices of the first occurence of the unique elements in same order
[~, idx] = unique(T.VerNo, 'stable')
%% Output
GsT = T(idx, :)
Note that the order of rows obtained above is according to the descending values of VerNo.
Plus de réponses (0)
Voir également
Catégories
En savoir plus sur Time Series Objects dans Help Center et File Exchange
Community Treasure Hunt
Find the treasures in MATLAB Central and discover how the community can help you!
Start Hunting!