Skip to main content

Documentation Index

Fetch the complete documentation index at: https://mintlify.com/backtest-kit/uzse-backtest-app/llms.txt

Use this file to discover all available pages before exploring further.

scripts/check-gaps.ts scans the raw trade-results collection in MongoDB and reports every calendar date between the first and last recorded trade that has no trading activity. Running it after importing new data and before building candles lets you confirm that your dataset is complete and that any gaps you find are genuine exchange closures rather than missing downloads.

​
Usage

No arguments are required. Run the script with tsx from the project root: 
npx tsx scripts/check-gaps.ts
The script connects to MongoDB using the MONGO_URI environment variable (default: mongodb://localhost:27017/backtest), performs its analysis, prints the results to stdout, and then disconnects cleanly.

​
What It Outputs

The script prints two lines of summary followed by a labeled list of every missing date: 
Trading days in DB: 247
Missing days (18):
  2023-08-12 (Сб)
  2023-08-14 (Пн)
  2023-08-15 (Вт)
  2023-08-16 (Ср)
  2023-08-17 (Чт)
  2023-08-18 (Пт)
  2023-08-19 (Сб)
  2023-08-20 (Вс)
  2023-08-21 (Пн)
  ...
  • Trading days in DB — the count of distinct calendar dates on which at least one trade record exists
  • Missing days — every date in the continuous range from the first to the last trading day that has no records, annotated with its Russian day-of-week abbreviation

​
Day-of-Week Labels

The script uses Russian language day-of-week abbreviations, matching the locale of the UZSE data source:
LabelDay
ВсSunday
ПнMonday
ВтTuesday
СрWednesday
ЧтThursday
ПтFriday
СбSaturday
A missing day labeled Вс (Sunday) or Сб (Saturday) is always expected — UZSE does not trade on weekends. Missing days on weekdays warrant closer investigation.

​
How It Works

The script follows a straightforward three-step process:
1

Load all trade timestamps

TradeModel.find({}, { time: 1 }) retrieves every document from the trade-results collection, projecting only the time field to keep memory usage low.
2

Extract unique trading dates

Each time value is converted to a YYYY-MM-DD key using a zero-padded formatter, and the results are deduplicated into a Set. This gives the exact set of calendar dates on which trading occurred.
3

Walk the date range and find gaps

The script iterates day-by-day from the earliest to the latest trading date. Any date not present in the set is pushed onto the missing array with its day-of-week label.
const DOW = ["Вс", "Пн", "Вт", "Ср", "Чт", "Пт", "Сб"];

for (let d = new Date(start); d <= end; d.setDate(d.getDate() + 1)) {
  const key = toKey(d);
  if (!days.has(key)) missing.push(`${key} (${DOW[d.getDay()]})`);
}

​
Why Gaps Occur

Not all gaps indicate missing data. Common reasons include:
  • Weekends — UZSE is closed every Saturday and Sunday; these will always appear as missing days
  • Public holidays — Uzbek national holidays (Nowruz, Independence Day, etc.) result in exchange closures
  • Exchange-ordered trading halts — The most notable example is the August 2023 Hamkorbank halt, during which trading was suspended for approximately ten days following the registration of a new share issuance that tripled the bank’s authorized capital. This type of gap reflects a deliberate corporate or regulatory event, not missing data
If you see a gap on a weekday that does not align with a known public holiday or announced trading halt, re-check your download scripts. The UZSE website paginates trade results and limits exports to one month at a time — it is possible that a page was missed during collection.

​
Gaps in trade-results vs. candle-items

Finding gaps in the trade-results collection does not mean your candle data will have gaps. The build-candles.ts script handles missing days automatically:
  • For any day with no trade records, it synthesizes a candle with open = high = low = close = last_close and volume = 0
  • This ensures the candle-items collection has a continuous, unbroken time series suitable for indicator calculations
check-gaps.ts is therefore a data validation tool, not a repair tool. Use it to understand your dataset and to provide context when interpreting indicator signals near gap periods — not to fix the candle output.

​
When to Run

Run check-gaps.ts in the following situations:
  • After importing new trade data — confirm that the newly added date range is complete and that no download pages were skipped
  • Before building candles — understand which gaps will be synthesized so you can contextualize the resulting candles correctly
  • When investigating unusual indicator signals — a spike or flat region in a Pine Script indicator near a known gap date is likely an artifact of the gap-fill candle rather than a genuine market signal

Applying Pine Script Indicators

Learn how gap-filled candles affect indicator calculations and how to interpret signals around non-trading periods.

Build docs developers (and LLMs) love

Get started for freeTalk to us