Country or region and language

    Choose your country or region and your language.

    Suggested for you

      All countries and regions

        Choose your character

        It shows in the top bar and on your profile. Change it any time.

        Interviews in R

        Work in the tool while you explain your decisions.

        Choose a role and level. Each interview has a new task, follow-up questions and a report that checks your work and your explanation.

        Speak or type. Sign in to save your results across devices.

        Timed assessment

        • 12 practical questions
        • 45 minutes
        • Mid level

        Questions are drawn from a bank of 36. Your report includes a skill breakdown and lessons to practice next.

        Practice interviews

        Junior tasks cover the fundamentals. Mid and senior tasks add debugging, changing requirements and review decisions.

        10 interviews

        • Data analyst
        • Junior
        • 30 min

        A signed depot revenue report

        Plover Dispatch needs result with one row per depot and net as the sum of amount. Keep credits and the depot with a missing label. Ignore missing amounts, then return only depot and net without grouping.

        Start interview
        Skills and rubricA signed depot revenue report
        • I group shipments by depot and sum signed amounts with missing amounts omitted because a credit reduces net revenue. The missing depot label stays as its own group. I drop grouping and reconcile net with the source sum before handing over the report.
        • The missing amount and the negative credit test different rules. I omit the unknown amount from this requested sum, but retain the recorded credit because removing it would overstate revenue. I also retain the unlabeled depot so its money remains visible.
        • I retain the source records, the saved work and the checked output because another analyst must be able to reproduce this result. I record the observation window and the unresolved limitation. The decision owner approves any exception before this result changes a business action.
        • Data scientist
        • Junior
        • 30 min

        A latest ticket revision repair

        Morrow Printworks receives ticket revisions out of order. Create result with exactly the highest revision for each ticket. Preserve ticket, revision and charge. Revision numbers are unique within a ticket.

        Start interview
        Skills and rubricA latest ticket revision repair
        • I sort each ticket by descending revision before retaining one row because arrival order does not establish which charge is current. I preserve all requested fields and compare the retained revision with the maximum source revision for every ticket.
        • Ticket a has more than one charge and arrives out of order. All-field deduplication preserves those revisions because their values differ. I use ticket as the identity and revision as precedence, then verify one output row per ticket.
        • I retain the source records, the saved work and the checked output because another analyst must be able to reproduce this result. I record the observation window and the unresolved limitation. The decision owner approves any exception before this result changes a business action.
        • Data analyst
        • Junior
        • 30 min

        A depot owner enrichment

        Lark Services needs result containing every shipment enriched with owner from directory on depot. Keep the shipment fields and owner. A missing depot is unknown and must never match the missing directory key.

        Start interview
        Skills and rubricA depot owner enrichment
        • I left-join on depot with missing keys forbidden from matching because an unknown depot does not identify the unknown directory record. Every shipment remains in result. I reconcile shipment identities and review blank owners as exceptions.
        • The shipment with a missing depot must remain, with a missing owner. Matching two missing values would invent an owner relationship because neither record supplies a depot identity. I check row count and unmatched identities separately.
        • I retain the source records, the saved work and the checked output because another analyst must be able to reproduce this result. I record the observation window and the unresolved limitation. The decision owner approves any exception before this result changes a business action.
        • Data scientist
        • Mid
        • 30 min

        A weekly allocation reshape

        Quarry Scheduling needs result in long form with depot, week and units. Turn WeekA and WeekB into rows and retain zero allocations. Keep only those three fields.

        Start interview
        Skills and rubricA weekly allocation reshape
        • Hill has a recorded zero in WeekA. I retain that record because deleting it would make zero allocation indistinguishable from a missing observation. The output needs both requested weeks for each depot.
        • I add the approved week column to the measures being pivoted and update the expected row count because a new identifier must not become a measure accidentally. I compare each depot and week with the source before refreshing the report.
        • I retain the source records, the saved work and the checked output because another analyst must be able to reproduce this result. I record the observation window and the unresolved limitation. The decision owner approves any exception before this result changes a business action.
        • Data analyst
        • Mid
        • 30 min

        A volume-weighted quality score

        Rowan Quality needs numeric result equal to the volume-weighted mean score. Use only rows with a recorded score and positive volume. An unknown score contributes neither score nor volume.

        Start interview
        Skills and rubricA volume-weighted quality score
        • The missing score has positive volume and the final score has zero volume. I exclude both from this weighted mean because one lacks the measured value and the other has no eligible weight. An unweighted mean answers a different question.
        • I return an explicit unavailable result and report zero eligible volume because dividing by zero cannot produce a meaningful score. I preserve the exception count and ask the owner whether the missing observation period should be held or annotated.
        • I retain the source records, the saved work and the checked output because another analyst must be able to reproduce this result. I record the observation window and the unresolved limitation. The decision owner approves any exception before this result changes a business action.
        • Data scientist
        • Mid
        • 30 min

        A February dispatch window

        Cove Dispatch supplies dated$stamp as day/month/year text. Parse it as a Date, retain February 1 through February 3, 2027 inclusive, and create result with id, posted and units in that order.

        Start interview
        Skills and rubricA February dispatch window
        • The strings 01/02 and 03/02 are ambiguous without the format contract. I treat them as February 1 and February 3 because the source specifies day/month/year. The January 31 and February 4 records test exclusion at each edge.
        • I confirm the business time zone and convert each timestamp before extracting the reporting date because a UTC day can differ from the local dispatch day. I test midnight on both boundaries and preserve invalid timestamps as exceptions.
        • I retain the source records, the saved work and the checked output because another analyst must be able to reproduce this result. I record the observation window and the unresolved limitation. The decision owner approves any exception before this result changes a business action.
        • Data analyst
        • Mid
        • 30 min

        A depot running balance

        Tern Inventory needs result with depot, seq, delta and balance. Sort by depot then seq ascending and calculate cumulative delta separately within each depot. Preserve signed adjustments and return an ungrouped table.

        Start interview
        Skills and rubricA depot running balance
        • Quay sequence two arrives before sequence one, while Hill includes a negative adjustment. I sort first and restart accumulation for each depot because sorting after cumsum would only rearrange already wrong balances.
        • I recompute that depot from the inserted sequence onward because each later balance depends on the earlier signed deltas. I preserve the other depot and reconcile both the changed final total and the unchanged prefix.
        • I retain the source records, the saved work and the checked output because another analyst must be able to reproduce this result. I record the observation window and the unresolved limitation. The decision owner approves any exception before this result changes a business action.
        • Data scientist
        • Senior
        • 30 min

        A dispatch demand regression

        Alder Planning needs numeric result with predictions for future$hours from an ordinary least-squares model of output on hours in training, including an intercept. Fit on training only and predict the three future rows in their existing order.

        Start interview
        Skills and rubricA dispatch demand regression
        • I remove or replace the late feature and rebuild validation because the planning decision cannot use a value recorded afterward. I keep a baseline based on information available at decision time and ask the planner to confirm that cutoff.
        • I retain the source records, the saved work and the checked output because another analyst must be able to reproduce this result. I record the observation window and the unresolved limitation. The decision owner approves any exception before this result changes a business action.
        • The future nine-hour observation exceeds the largest training hour value. I can compute the requested linear prediction, but I flag extrapolation because the fitted relationship has no observations there. Reversing the response and predictor fits a different loss.
        • Data analyst
        • Senior
        • 30 min

        A safe utilization function

        Reed Resourcing needs safe_rate(numerator, denominator). Return numerator divided by denominator, but return NA_real_ where the denominator is zero or missing. Preserve vector length and missing numerators. The checker calls your function on unseen vectors.

        Start interview
        Skills and rubricA safe utilization function
        • I add a negative-denominator validation rule and report those source rows because a negative capacity is now an input error. I retain the existing zero and missing policy, then test the revised function on mixed vectors before releasing it.
        • I retain the source records, the saved work and the checked output because another analyst must be able to reproduce this result. I record the observation window and the unresolved limitation. The decision owner approves any exception before this result changes a business action.
        • A zero numerator over a positive denominator is zero, while a zero denominator is unavailable. I keep those cases separate because replacing both with zero would hide missing capacity. A missing numerator stays missing even with a valid denominator.
        • Data scientist
        • Senior
        • 30 min

        A missing-week completion

        Brookline Scheduling supplies allocations with depot, week and units. Create result with every combination of the observed depots and weeks 1, 2 and 3. Fill only newly absent combinations with zero units; preserve an existing missing units value.

        Start interview
        Skills and rubricA missing-week completion
        • I need the authoritative depot roster because observed depots alone cannot reveal a depot absent from the file. I complete against that approved roster and the requested weeks, then retain the existing distinction between an unknown value and an absent allocation.
        • I retain the source records, the saved work and the checked output because another analyst must be able to reproduce this result. I record the observation window and the unresolved limitation. The decision owner approves any exception before this result changes a business action.
        • The existing missing units value differs from the absent week row. I preserve the unknown measurement and fill only newly created combinations because the business rule assigns zero to absence, not to every missing value.

        Interview Hub access

        The Interview Hub is included in All Access. Choose a plan or sign in to your account to continue.

        View plansSign in or join

        Your Work Gym account

        Save your progress and pick up on another device. Your browser's progress comes with you.

        or

        You can keep practicing without an account.

        Change your plan

        Checking the price

        Help and feedback

        Ask a question, tell us what is not working, or share an idea. We reply by email.

        What is it about?

        We use it only to reply to you.