Skip to content

Test Suite

Jump to bottom
crux1s edited this page Mar 27, 2026 · 3 revisions

Test Suite

test_tastymechanics.py contains 299 tests across 24 sections. Tests run directly against real CSV data with no Streamlit server required.

Running the tests

PYTHONIOENCODING=utf-8 python3 test_tastymechanics.py

Place your TastyTrade CSV (named tastytrade_*.csv or tastymechanics_*.csv) in the same folder as the script. The suite will find it automatically.

A passing run looks like:

GRAND TOTAL:  299 tests  |  299 passed  |  0 failed
All tests passed ✅

A failing run exits with code 1 and prints every failed test with the actual vs expected value.

What is tested

Section Coverage
1. Data loading & parsing Row counts, column types, date parsing
2. FIFO engine Per-ticker realised gains, fractional shares, lot matching
3. Options cash flows Total options P&L, SPX settlement, futures
4. Dividends & interest Income component totals
5. Campaign accounting Window boundary verification — pre vs post-purchase option attribution
6. Total realized P&L All three components summed correctly
7. Deposits / portfolio stats Deposits, withdrawals, ROR, cash balance
8. Open equity positions Share counts, cost basis, blended basis
9. Windowed P&L consistency Consistency check against ground truth
10. Edge cases Assignments, expirations, cash-settled index, transfers
11. Individual campaign cards Per-ticker premiums, effective basis, campaign P&L
12. Outside-window options Pre-purchase standalone option P&L
13. Windowed P&L — named windows YTD, Last 7 Days, Last Month, Last 3 Months, All Time values
14. Capital deployed Per-ticker and total capital in shares
15. Ticker-level options P&L Per-ticker option cash flows
16. Self-calibrating invariants Structural checks that work on any CSV
17. Closed trades — core Total count, win rate, P&L, per-ticker, spot trades
18. Closed trades — strategy Count and P&L per strategy type
19. Closed trades — close types Expired/assigned/closed counts, debit trades
20. Closed trades — windows YTD/7d/30d filtering, boundary checks
21. Union-Find helpers _uf_find, _uf_union, _group_symbols_by_order
22. calc_dte DTE calculation, edge cases, non-option rows
23. build_option_chains Chain detection, rolls, multi-type, empty input
24. UI helpers xe() HTML escaping, identify_pos_type(), detect_strategy()

VERIFIED tests

Four tests are marked VERIFIED — the exact values were cross-checked screenshot-by-screenshot against the live TastyTrade transaction history UI:

  • SLV Put — Jan 7 → Jan 10 2026, credit $100.88, net P&L $39.76
  • INTC Losing Put — Jan 28 → Feb 17 2026, credit $103.88, net P&L -$13.24
  • SMR Losing Put — Jan 26 → Feb 17 2026, credit $60.88, net P&L -$87.24
  • TSLA Call Debit Spread — Oct 2 BTO 457.5C/$12.23 + STO 460C/$10.93, both expired worthless Oct 4. Net -$132.24, Capture % = None (not -100%)

These tests confirm trade pairing logic, debit trade display, and that Capture % is correctly suppressed for non-credit trades.

Section 16 — Self-calibrating invariants

Section 16 contains tests that are valid for any CSV — they check structural correctness rather than specific values:

  • All expiration rows sum to $0
  • All assignment option rows sum to $0
  • All equity buy rows have positive Net_Qty_Row
  • All equity sell rows have negative Net_Qty_Row
  • FIFO equity P&L is less than or equal to gross sale proceeds
  • All Time window P&L equals the sum of components
  • Net deposited is >= 0
  • Total capital deployed is > 0

These never need updating and will catch structural bugs on any account's CSV.

Section 24 — UI helpers

Three functions tested in Section 24:

xe(value) — HTML-escapes strings passed into rendered HTML. Tests confirm <, >, &, " are all escaped correctly, non-string types pass through unchanged, and None returns None.

identify_pos_type(row) — classifies a position row as Long/Short Stock, Long/Short Call, Long/Short Put, Future Option Short Put, or Asset. Tests cover all branches including unknown instrument types.

detect_strategy(df) — classifies a group of legs into a strategy name. Tests cover: Short Put, Covered Call, Covered Strangle, Short Strangle, Jade Lizard, Big Lizard, Risk Reversal, Call Butterfly, Put Butterfly, Calendar Spread, Call Debit Spread, Long Call, Long Stock, and Custom/Mixed fallback.

Two bugs were found and fixed during this section:

  • Call Butterfly false positive — was incorrectly matching Call Debit Spread when lc==2, sc==1 with 3 strikes. Fixed by verifying the short call is the middle strike.
  • Long Call false positive — lc>0 matched groups with 2+ unmatched long calls. Fixed to lc==1.

Updating tests after a new CSV export

When you export a fresh CSV with new trades, some aggregate tests will fail because the snapshot values have changed. To update:

  1. Run the tests and note which fail
  2. The failures will be aggregate counts and totals (row count, total P&L, window P&L etc.)
  3. Update the expected values in the relevant check() calls
  4. The VERIFIED tests and invariant tests (section 16) will not need updating — they are structural

Adding new tests

When adding a new feature or fixing a bug, add tests at the same time:

  1. Find the real values from your CSV using a quick Python snippet
  2. Add check() or check_int() calls in the appropriate section
  3. If you manually verified a value against the TastyTrade UI, prefix the test name with VERIFIED
  4. Run the full suite to confirm all 299 (+ your new ones) pass

Clone this wiki locally