Practising

The loop#

Every problem goes through the same commands, whatever the language:

cstudy next  ->  cstudy start <slug> --lang kotlin  ->  edit in your editor  ->  cstudy test --watch
         ->  cstudy bench (optional)  ->  cstudy done --rating ok  ->  cstudy solution (optional)

Commands that take a problem accept its slug (two-sum), its id, or the path of one of your solution files. Without one, they act on the active attempt: your most recent attempt still in progress.

Choosing a language#

Pass --lang python, typescript, java, kotlin, go or rust to cstudy start, or set default_lang in your configuration. Reviews are scheduled per problem, so you can solve a problem in one language and review it in another.

A file path names both the problem and the language, which is how editors run cstudy on the open file: cstudy test python/practice/two_sum.py is cstudy test two-sum --lang python.

Starting and resuming#

cstudy start <slug> writes the stub, renders the statement to .statements/<slug>.md, begins an attempt and opens both files in $VISUAL or $EDITOR. Starting a problem you already have in progress in the same language resumes that attempt and keeps your file.

Set open_on_start = false in your configuration if you never want cstudy to launch the editor.

Testing#

cstudy test runs your code against the problem’s cases. With --watch it runs again each time you save. Visible cases are always listed; passing hidden cases collapse into a count unless you pass --all; a failing hidden case shows its input, the expected answer and yours.

$ cstudy test --time-limit
two-sum · kotlin · 14 cases (2 visible, 11 hidden, 1 generated) · time limit 5x · timer 12:31 / 25:00
build  1.8s (gradle, warm)

PASS     ex1   0.04ms
PASS     ex2   0.03ms
FAIL     h3    hidden
         input     nums=[-3, 8, 0, 3]  target=0
         expected  [0, 3]  (unordered)
         actual    [3, 3]
TLE      g1    hidden · 412ms > limit 95ms (5x hash-map, calibrated)
PASS     10 hidden cases

12/14 passed · 1 failed · 1 tle · total 2.6s

Useful flags

On a full pass the summary ends with all passed. Next: cstudy done --rating easy|ok|hard. Every run except a debug run is recorded, so cstudy knows whether your latest run passed when you finish.

Hints and solutions#

cstudy hint reveals the next hint of the problem. Hints are counted per attempt, and using any caps the grade at ok.

$ cstudy hint
two-sum · hint 1 of 2
For each value, which partner would complete the sum?

cstudy solution shows, through your pager, the explanation of the optimal approach and its code in your language. Before your attempt has passed, it asks first, because revealing the solution grades the attempt as again:

Reveal the solution? The attempt will be graded as again. [y/N]

Finishing an attempt#

cstudy done closes the attempt and schedules the review. When your latest test run passed every case, rate how it went with --rating easy, ok or hard; when it did not, finish with --gave-up. Add --notes "…" to keep a note on the attempt.

$ cstudy done --rating ok
two-sum · kotlin · passed · 18m 40s · rating ok · hints 0
next review  2026-10-27 (in 3 days) · ease 2.50

The rating becomes a grade, and some things cap it. The lowest cap wins:

Solve time is the wall-clock time from cstudy start to cstudy done; there is no pause.

Reviews and spaced repetition#

cstudy schedules each problem with a simplified SM-2 algorithm. A good grade pushes the next review further out; a lapse brings the problem back tomorrow. This is how one problem moves through 6 reviews:

Intervals never exceed 180 days (max_interval_days). A problem counts as mastered once its interval reaches 21 days with a last grade of ok or better, and as due on and after its review date.

What cstudy next picks#

cstudy next answers “what should I practise now?” the same way every time for the same progress and date:

1. A due review, ranked by how overdue it is relative to its interval, plus how weak its pattern is.
2. Otherwise a new problem from the weakest pattern (or, outside the algorithms track, the weakest track). Within it, easy problems come first; medium opens once you pass an easy one, and hard once you pass a medium one.
3. Otherwise it prints nothing due and no new problems in scope.

$ cstudy next
review  merge-intervals  (medium · intervals)  overdue 3d, interval 7d
start   cstudy start merge-intervals --lang kotlin

$ cstudy next
new     koko-eating-bananas  (medium · binary-search)  weakest pattern: binary-search (0.75)
start   cstudy start koko-eating-bananas --lang kotlin

Weakness combines your recent fail rate, how much of the group you have mastered, and how long ago you last practised it. An untouched group scores 0.75, so cstudy rotates through new patterns before going deep on one.

Checking performance#

cstudy bench runs your solution and every reference approach on the same generated inputs at growing sizes, fits how the time grows, and compares you with the optimal reference. It checks your answers at every size too.

$ cstudy bench
two-sum · python · sizes 1e2..1e5 · reps 7 · power AC · load 1.2

TARGET        DECLARED   n=1e2    n=1e3    n=1e4    n=1e5    SLOPE  BAND     MEMORY
brute-force   O(n^2)     0.21ms   19.8ms   2.04s    budget   2.00   ~n^2     constant
hash-map      O(n)       0.01ms   0.09ms   0.94ms   10.1ms   1.00   ~n|nlogn linear
you                      0.01ms   0.11ms   1.10ms   11.8ms   1.01   ~n|nlogn linear

verdict   optimal: matches hash-map, 1.17x at n=1e5 · confidence high (spread 3%, R^2 0.999)
space     linear, same as hash-map · confidence high
history   best previous 13.0ms at n=1e5 (2026-09-12) · now 9% faster · 1st of 3 runs

Each verdict comes with a confidence, high, medium or low, and a warning when your machine was busy or on battery. Benchmarks compare you with the references and with your own earlier runs, never with other people.

Timers and mock interviews#

Any attempt can have a time box: cstudy start two-sum --timer 25m. cstudy test then shows the time left, and finishing over it caps the grade at hard.

cstudy mock simulates an interview. It picks new medium and hard problems, each from a different group where it can, starts them together with one shared timer, and locks cstudy hint and cstudy solution until the timer ends or every problem is done.

$ cstudy mock
mock interview · 2 problems · 45 min shared timer · python
1  group-anagrams  (medium · arrays-hashing)  weakest pattern: arrays-hashing (0.75)
   stub       python/practice/group_anagrams.py
   statement  .statements/group-anagrams.md
2  lru-cache  (medium · linked-list)  weakest pattern: linked-list (0.75)
   stub       python/practice/lru_cache.py
   statement  .statements/lru-cache.md
locked     cstudy hint and cstudy solution, until the timer ends or every problem is done
next       cstudy test group-anagrams --watch

In a mock attempt of a project problem, cstudy submit reports only how many hidden cases failed per level, as an interview would.

Tracking progress#

cstudy stats summarises your progress per group: problems, attempts, passes, mastered problems, the pass rate of the last 10 attempts, median solve time and weakness. Below come your streak, the reviews due and your latest benchmark verdicts.

$ cstudy stats
GROUP           PROBLEMS  ATTEMPTED  PASSED  MASTERED  PASS RATE  SOLVE TIME  WEAKNESS
arrays-hashing        17          5       4         1        80%     14m 10s      0.42
two-pointers          10          0       0         0          -           -      0.75
system-design         10          1       1         0       100%     45m 00s      0.30

attempts  6 finished · 5 in the last 7 days · streak 3 days
reviews   1 due today · 1 overdue
bench     4 benched · 2 optimal · 1 near-optimal · 1 suboptimal · 0 unknown

Group by --by pattern (the default), track, lang or difficulty; narrow the attempt columns with --since 2026-10-01; or print --json. To browse problems with their status and next review date, use cstudy list, filtered with --status new|attempted|passed|due|mastered:

$ cstudy list
ID    SLUG              DIFF    PATTERN          STATUS    DUE         REFS
0001  two-sum           easy    arrays-hashing   passed    2026-10-27  py ts java kt
0002  lru-cache         medium  linked-list      new       -           py ts java kt
0014  koko-eating-...   medium  binary-search    due       2026-10-21  py java