TDICDTS, the Qlik Talend Data Integration Developer exam, names the tMap component in four of its 42 objectives, and a fifth (the filter reject output) is configured inside tMap too. If you can join two sources in tMap, catch the rows that fail the join, and split the output with filters and a reject table, you have covered most of the "Working with files" and "Joining and filtering data" areas in one skill.
This entry walks through that skill the way the exam lists it. The exam facts were read on Qlik's own pages on 20 November 2026, and the tMap behaviour comes from the Talend Studio documentation on help.qlik.com.
What is the TDICDTS exam and what does it test?
TDICDTS is the exam for the Talend Data Integration Certified Developer using Talend Studio badge. It tests whether you can build, test and run data integration Jobs in Talend Studio, alone and in a team project.
The official TDICDTS exam details page gives the format:
- 55 questions in 90 minutes
- Passing score: 70%
- Qlik's note: the number and difficulty of questions may change, and the passing score is adjusted to keep a consistent standard
- Recommended: at least six months of experience with Talend products
- Price: $250 USD, paid by card through Qlik Learning
- All questions are multiple choice; when more than one answer is needed, the question says how many
- English only; non-native speakers can ask Qlik support for extra time
The topic list has 10 areas and 42 objectives. Qlik does not publish a weight for each area, so the counts below are objectives, not exam percentages.
Where does tMap appear in the TDICDTS objectives?
It sits in two neighbouring areas, which together hold 9 of the 42 objectives (about 21% of the list).
"Working with files" has four objectives:
- Configure basic component properties
- Create and configure a schema
- Use the tMap component and configure a simple mapping
- Use pre-defined Talend Java functions
"Joining and filtering data" has five:
- Define and configure Talend metadata
- Join two sources of data using the tMap component
- Define the tMap Join settings and its reject capture mechanism
- Create a filter condition in tMap
- Configure a filter reject output along with multiple filtered outputs
Read the verbs: use, join, define, create, configure. These are build tasks, not definitions. A candidate who has only read about tMap tends to struggle with questions that describe a Job and ask which setting produces a given output.
How does a tMap join work in Talend Studio?
A tMap join links a main flow to one or more lookup flows on a key, and each output table takes columns from both.
In the Map Editor you drag a column from the main input onto the matching column of the lookup. Talend's documentation on using an Inner Join in the tMap editor describes what happens next: a purple link appears, and Studio creates a key that is used as a hash key to speed up the match search.
Three settings on the lookup then decide the result.
Join model: Inner Join or Left Outer Join
With Left Outer Join, a main row that finds no match in the lookup still goes to the output, with the lookup columns empty. With Inner Join, that row does not reach the main output. The documentation puts it as: Inner Join "avoids that null values are passed on to the main output flow."
Match model: Unique, First or All matches
When the lookup holds more than one row for the same key, the match model picks what comes out. Talend's example Job states that Unique match is the default and outputs the last matching lookup record. First match outputs the first matching record, and All matches outputs every matching record, so one main row can become several output rows.
Lookup model: when the lookup is loaded
Load once is the default: the lookup is read once, before the main rows run. Reload at each row reads it again for every main row, and the cached version only loads rows it does not already hold. This matters when the lookup is large, or when its query depends on a value from the main row.
What happens to the rows an Inner Join rejects?
They are dropped unless you add an output to catch them. That catch is the "reject capture mechanism" in the objective.
You add a new output table in tMap, open its settings, and set Catch lookup inner join reject to true. Every main row that failed the Inner Join now lands in that table. The Studio guide adds a rule worth remembering: an Inner Join table should always be coupled to an Inner Join Reject table. In real Jobs this is how you find bad keys instead of losing rows without a trace.
Talend's component documentation has a full worked scenario on mapping data with inner join rejections in tMap. A Cars file is joined to a Resellers lookup on the reseller ID. Matched cars go to one output, and cars with an unknown reseller go to a reject output. When there are two lookups, a filter such as "Resellers_data.ID_Reseller==null" on each reject table shows which join failed. Build that scenario yourself once and most join questions become easy to picture.
How do tMap filters and the filter reject output work?
A filter is an expression on an output table; only rows for which it is true go to that table. A table with Catch output reject set to true collects the rows that passed no filter.
Talend's guide describes the reject table as the ELSE of all the regular tables' filters. Three details from the documentation tend to decide exam answers:
- Regular (non-reject) tables are checked first; the reject tables come after.
- A row is not routed to only one output. A row that meets one table's filter is still checked against the others and can land in several outputs.
- You can have more than one reject table, and you tell them apart with their own filter lines.
So "multiple filtered outputs" in the objective means several tables with their own expressions, and the "filter reject output" is the table that gets whatever none of them took.
What does one worked example look like?
Here is a small Job to build in Studio. It covers every tMap objective in one go.
- Main input: an orders file with order ID, customer ID, country code and amount.
- Lookup: a customers file with customer ID and customer name.
- Join on customer ID as an Inner Join, match model Unique match.
- Output 1, "domestic": filter on the country code for your own country.
- Output 2, "large_export": filter where the country is not yours and the amount is above a threshold.
- Output 3, "no_route": Catch output reject = true.
- Output 4, "unknown_customer": Catch lookup inner join reject = true.
- In the domestic output, use a pre-defined Talend Java function to upper-case the customer name.
Now change one thing at a time and predict the row counts before you run: switch to Left Outer Join, add a duplicate customer to the lookup and try First match and All matches, then give an order a country that fits both filters. Writing down your prediction first is what turns this into exam practice.
If you want to see how a different exam family lays out its data-modelling work, my earlier entry on the Qlik Sense Data Architect (QSDA) exam changes covers the Qlik Sense side of Qlik's catalogue.
How should you study the tMap part in four weeks?
This plan assumes you already have Talend Studio installed and a few months of hands-on use.
- Week 1: files and schemas. Read and write delimited files, build a schema by hand and from metadata, and do a simple tMap mapping with expressions and Java functions.
- Week 2: joins. Build the Cars and Resellers scenario, then the orders Job above. Test each join model, match model and lookup model, and note the row counts.
- Week 3: filters and rejects. Add several filtered outputs, a Catch output reject table, and two reject tables told apart by filters. Check which rows land in more than one output.
- Week 4: connect it to the rest of the exam. Put context variables into the file paths, wrap the Job in a master Job, and read the official topic list line by line, marking every objective you have actually built.
The tMap areas are only 9 of 42 objectives, so keep the other eight areas moving in parallel.
Studying more than one data exam this season? The same objective-by-objective method works for data science too; see these notes on the Dell D-DS-FN-23 data science exam by topic weight.
Which tMap mistakes cost points?
- Assuming one row goes to one output. Talend's own guide says a row is checked against every filter and can reach several outputs.
- Mixing up the two reject settings. Catch lookup inner join reject is for failed joins; Catch output reject is for rows no filter accepted.
- Forgetting the default match model. Unique match is the default and returns the last matching lookup record, not the first.
- Ignoring duplicates in the lookup. All matches can multiply rows; a question about "more output rows than input rows" usually points here.
- Leaving an Inner Join without a reject table. It works, but the guide says the two should always be coupled.
What else should you know before booking TDICDTS?
The exam is open book: Qlik allows its learning courses, the Qlik and Talend products, the documentation, knowledge-base articles and your own notes. AI or AI-assisted tools and web browsers are strictly prohibited, and the session is recorded and proctored online through Webassessor. Open book does not mean slow book: 55 questions in 90 minutes is under 100 seconds each, so know where the tMap pages are in the help center before you start.
After payment, the exam is available within 24 business hours, and there is no fixed date to book. Qlik's page on certificates, badges and retake rules covers the rest: the result shows at once, the Credly badge invitation follows within 24 to 48 hours, and there are no paper certificates. You get three attempts, each paid in full, with 48 hours after the first fail and 14 calendar days after the second. The certification is valid for two years.
For the full objective list next to sample questions and a summary of the format, the TDICDTS syllabus and exam summary on AnalyticsExam puts everything on one page.
Frequently asked questions
1. How many tMap objectives are on the TDICDTS exam?
Four of the 42 objectives name tMap: a simple mapping, joining two sources, the join settings and reject capture, and a filter condition. A fifth, the filter reject output, is also configured in tMap.
2. What is the passing score for TDICDTS?
The passing score is 70%. Qlik notes that it adjusts the passing score to keep a consistent standard when exam content changes.
3. How long is the TDICDTS exam and what does it cost?
It has 55 multiple-choice questions in 90 minutes and costs $250 USD, paid by card through Qlik Learning.
4. Can I use the Talend documentation during the exam?
Yes. Qlik exams are open book, so the documentation, knowledge-base articles and your own notes are allowed. AI or AI-assisted tools and web browsers are prohibited.
5. How many times can I retake TDICDTS?
You get three attempts, each paid in full. You wait 48 hours after the first fail and 14 calendar days after the second.

