Expression Rules and Syntax

1. Use Cases

Expressions are BeeQuant's core language for describing factor calculations, trading conditions, and label definitions. They are primarily used in the following places:

Location	Purpose	Example	Supported features
Data-processing node	Factor expression	`rsi = ta_rsi(close, 14)`	Operator functions + operators
	Label expression	`label = shift(close, -5) / shift(open, -1) - 1`	Operator functions + operators
	Filter expression	`volume > 1000000`	Comparison + logical operations
Strategy-type node	Timing entry/exit expression	`(rsi > 70) & (ma5 > ma20)`	Comparisons + logical operations only
Strategy-type node	Asset-selection filter expression	`volume > 1000000`	Comparisons + logical operations only

⚠️ Note: Condition expressions in strategy-type nodes—timing entries/exits and asset-selection filters—cannot call time-series operator functions such as ta_ma, shift, or ta_cross_over. Compute every factor in the data-processing node first, then use only the emitted column names in comparisons and logical operations.

2. Basic Syntax

An expression consists of column names, constants, operators, and function calls.

2.1 Built-in Columns

The data-processing and extraction node retrieves the following columns from the exchange. You can reference them directly in its factor, label, and filter expressions:

Column	Description	Example
`open`	Opening price	`(close - open) / open`
`high`	Highest price	`high - low`
`low`	Lowest price	`close - low`
`close`	Closing price (most commonly used)	`ta_ma(close, 20)`

⚠️ Reference ≠ output: open/high/low/close/volume and fundamental fields may participate in every expression field of the data-processing and extraction node (factor / label / filter), but they are not forwarded downstream by default. Its output contains only timestamp, symbol, and assigned variables whose names do not begin with _. A strategy condition can reference only columns actually emitted upstream (factor, pass-through, or AI prediction columns). To use close in a strategy condition, explicitly pass it through in a factor expression, for example close = close.

2.2 Constants

Type	Description	Examples
Integer	Used for window lengths, period parameters, and similar values	`14`, `20`, `60`
Floating-point number	Used for thresholds, ratios, and similar values	`0.05`, `2.0`, `-0.5`
Boolean	True/false	`True / 1`, `False / 0`

2.3 Variable Definitions

In factor configuration, define variables as variable_name = expression. Put each variable on a separate line.

Syntax

# Format: variable_name = expression
_ma20 = ta_ma(close, 20)// Intermediate variable; not output
diff = _ma20 - close// Output variable (difference between the moving average and closing price)
rsi = ta_rsi(close, 14)// Output variable

Example output from the configuration above (_ma20 is an intermediate variable and is omitted; diff and rsi are output):

timestamp	symbol	close	diff	rsi
2024-12-11 00:00:00	BTCUSDT	97800.0	-679.5	58.32
2024-12-11 01:00:00	BTCUSDT	97750.0	-593.2	55.18
2024-12-11 02:00:00	BTCUSDT	97820.0	-616.8	59.45
2024-12-11 03:00:00	BTCUSDT	98050.0	-781.5	63.21

● Gold columns are newly added output columns. _ma20 is omitted because its name begins with _.

💡 When to use intermediate variables:They make complex factor calculations easier to read.

# Example: calculate the position within Bollinger Bands
_ma20 = ta_ma(close, 20)// Intermediate variable; not output
_std = t_std(close, 20)// Intermediate variable; not output
_upper = _ma20 + 2 * _std// Intermediate variable; not output
_lower = _ma20 - 2 * _std// Intermediate variable; not output
bb_pos = (close - _lower) / (_upper - _lower)// Final factor; output

3. Operators

3.1 Arithmetic Operators

Symbol	Name	Example	Description
+	Addition	`close + open`	Adds two values
-	Subtraction	`high - low`	Subtracts one value from another
*	Multiplication	`close * volume`	Multiplies two values
/	Division	`(close - open) / open`	Divides one value by another
**	Exponentiation	`returns ** 2`	Raises x to the power y
%	Modulo	`index % 5`	Returns the remainder after division

3.2 Comparison Operators

Comparison operators return Boolean values and are used primarily in strategy-node condition expressions.

Symbol	Name	Example	Description
>	Greater than	`rsi > 70`	RSI is overbought
>=	Greater than or equal to	`close >= ma20`	Price is at or above the moving average
<	Less than	`rsi < 30`	RSI is oversold
<=	Less than or equal to	`close <= ma20`	Price is at or below the moving average
==	Equal to	`signal == 1`	Signal equals 1
!=	Not equal to	`trend != 0`	Trend is not 0

3.3 Logical Operators

Logical operators combine multiple conditions.

Symbol	Name	Example	Description
&	AND	`(rsi < 30) & (close > ma20)`	Both conditions must be true
\|	OR	`(rsi > 70) \| (rsi < 30)`	Either condition may be true
~	NOT	`~(close > open)`	Negates the condition

⚠️ Note: When using & or |, wrap each condition in parentheses to avoid operator-precedence issues.

Operator precedence (highest to lowest)

1. ** Exponentiation

2. ~ Logical NOT

3. * / % Multiplication, division, modulo

4. + - Addition and subtraction

5. > >= < <= Comparisons

6. == != Equality tests

7. & Logical AND

8. | Logical OR

💡 Use parentheses to make the order of operations explicit and improve readability.

4. Function Calls

Function-call syntax is function_name(argument1, argument2, ...). Calls may be nested. See Expression Operators for the complete function list.

Common function categories

Moving averages

ta_ma, ta_ema, ta_wma, t_mean

Momentum indicators

ta_rsi, ta_macd, ta_kdj, ta_cci

Volatility

ta_atr, ta_bbands, t_std

Time shifts

shift, pct, log, diff

Conditional functions

if, max, min, abs

Cross-sectional functions

c_rank, c_pctrank, c_zscore

Function-call examples

# Simple call
ta_ma(close, 20)// 20-period moving average

# Nested call
t_mean(ta_rsi(close, 14), 20)// 20-period average of RSI

# Conditional function
if(close > open, 1, -1)// Returns 1 for a bullish candle and -1 for a bearish candle

⚠️ Note: Time-series operator functions such as ta_ma, shift, and ta_cross_over may be used only in factor and label expressions in the data-processing node. Strategy condition expressions cannot call time-series functions; they may only compare and combine precomputed columns.

5. Factor Expression Examples

The following expressions are commonly used in the factor configuration of a data-processing node:

Factor	Expression	Description
Moving average	`ma20 = ta_ma(close, 20)`	20-period simple moving average
RSI	`rsi = ta_rsi(close, 14)`	14-period Relative Strength Index
Return	`ret = pct(close, 1)`	Single-period return
Volatility	`vol = t_std(pct(close, 1), 20)`	20-period standard deviation of returns
Volume ratio	`vol_ratio = volume / t_mean(volume, 20)`	Volume divided by 20-period average volume
Momentum	`mom = close / shift(close, 20) - 1`	20-period price rate of change
Bollinger position	`bb_pos = (close - ta_bbands_lower(close, 20, 2)) / (ta_bbands_upper(close, 20, 2) - ta_bbands_lower(close, 20, 2))`	Price position within the Bollinger Bands (0–1)

6. Condition Expression Examples

The following expressions are commonly used as entry and exit conditions in strategy-type nodes. A condition expression may directly reference factor variables defined by the upstream data-processing node.

🚫 Important: A strategy condition cannot call operator functions such as ta_ma, shift, or ta_cross_over. Compute every factor in the data-processing node first. Strategy conditions support only comparisons and logical operations on column names.

Condition	Strategy condition expression	Prerequisite in data processing
RSI overbought	`rsi > 70`	`rsi = ta_rsi(close, 14)`
RSI oversold	`rsi < 30`	`rsi = ta_rsi(close, 14)`
Golden cross signal	`golden == True`	`golden = ta_cross_over(ma5, ma20)`
Death cross signal	`death == True`	`death = ta_cross_under(ma5, ma20)`
Break above upper band	`close > bb_upper`	`bb_upper = ta_bbands_upper(close, 20, 2)`
Price and volume breakout	`(close > open) & (vol_ratio > 1.5)`	`vol_ratio = volume / t_mean(volume, 20)`
Multiple conditions	`(rsi < 30) & (close > ma20) & (vol_ratio > 1)`	`rsi / ma20 / vol_ratio are all precomputed`
AI prediction signal	`pred > 0.6`	`pred is emitted by the AI training node`

💡 Tip: A Boolean factor column such as golden or death may be referenced directly (golden) or compared explicitly (golden == True). Both forms have the same effect.

7. Label Expression Examples

The following expressions are commonly used for label definitions in a data-processing node. A label defines an AI model's prediction target and usually uses shift() to access future data.

Label	Expression	Task type
Future return	`label = shift(close, -5) / shift(open, -1) - 1`	Regression / binary classification
Future maximum gain	`label = shift(t_max(high, 5), -5) / shift(open, -1) - 1`	Regression
Cross-sectional ranking	`label = shift(close, -6) / shift(open, -1) - 1`	Ranking (converted by the engine)

💡 About shift():shift(col, n) shifts data backward when n > 0 or forward when n < 0. For example, shift(close, -5) means “the closing price five periods later.”

Expression Rules and Syntax

1. Use Cases

Expressions are BeeQuant's core language for describing factor calculations, trading conditions, and label definitions. They are primarily used in the following places:

Location	Purpose	Example	Supported features
Data-processing node	Factor expression	`rsi = ta_rsi(close, 14)`	Operator functions + operators
	Label expression	`label = shift(close, -5) / shift(open, -1) - 1`	Operator functions + operators
	Filter expression	`volume > 1000000`	Comparison + logical operations
Strategy-type node	Timing entry/exit expression	`(rsi > 70) & (ma5 > ma20)`	Comparisons + logical operations only
Strategy-type node	Asset-selection filter expression	`volume > 1000000`	Comparisons + logical operations only

2. Basic Syntax

An expression consists of column names, constants, operators, and function calls.

2.1 Built-in Columns

The data-processing and extraction node retrieves the following columns from the exchange. You can reference them directly in its factor, label, and filter expressions:

Column	Description	Example
`open`	Opening price	`(close - open) / open`
`high`	Highest price	`high - low`
`low`	Lowest price	`close - low`
`close`	Closing price (most commonly used)	`ta_ma(close, 20)`

2.2 Constants

Type	Description	Examples
Integer	Used for window lengths, period parameters, and similar values	`14`, `20`, `60`
Floating-point number	Used for thresholds, ratios, and similar values	`0.05`, `2.0`, `-0.5`
Boolean	True/false	`True / 1`, `False / 0`

2.3 Variable Definitions

In factor configuration, define variables as variable_name = expression. Put each variable on a separate line.

Syntax

Example output from the configuration above (_ma20 is an intermediate variable and is omitted; diff and rsi are output):

timestamp	symbol	close	diff	rsi
2024-12-11 00:00:00	BTCUSDT	97800.0	-679.5	58.32
2024-12-11 01:00:00	BTCUSDT	97750.0	-593.2	55.18
2024-12-11 02:00:00	BTCUSDT	97820.0	-616.8	59.45
2024-12-11 03:00:00	BTCUSDT	98050.0	-781.5	63.21

● Gold columns are newly added output columns. _ma20 is omitted because its name begins with _.

💡 When to use intermediate variables:They make complex factor calculations easier to read.

3. Operators

3.1 Arithmetic Operators

Symbol	Name	Example	Description
+	Addition	`close + open`	Adds two values
-	Subtraction	`high - low`	Subtracts one value from another
*	Multiplication	`close * volume`	Multiplies two values
/	Division	`(close - open) / open`	Divides one value by another
**	Exponentiation	`returns ** 2`	Raises x to the power y
%	Modulo	`index % 5`	Returns the remainder after division

3.2 Comparison Operators

Comparison operators return Boolean values and are used primarily in strategy-node condition expressions.

Symbol	Name	Example	Description
>	Greater than	`rsi > 70`	RSI is overbought
>=	Greater than or equal to	`close >= ma20`	Price is at or above the moving average
<	Less than	`rsi < 30`	RSI is oversold
<=	Less than or equal to	`close <= ma20`	Price is at or below the moving average
==	Equal to	`signal == 1`	Signal equals 1
!=	Not equal to	`trend != 0`	Trend is not 0

3.3 Logical Operators

Logical operators combine multiple conditions.

Symbol	Name	Example	Description
&	AND	`(rsi < 30) & (close > ma20)`	Both conditions must be true
\|	OR	`(rsi > 70) \| (rsi < 30)`	Either condition may be true
~	NOT	`~(close > open)`	Negates the condition

⚠️ Note: When using & or |, wrap each condition in parentheses to avoid operator-precedence issues.

Operator precedence (highest to lowest)

1. ** Exponentiation

2. ~ Logical NOT

3. * / % Multiplication, division, modulo

4. + - Addition and subtraction

5. > >= < <= Comparisons

6. == != Equality tests

7. & Logical AND

8. | Logical OR

💡 Use parentheses to make the order of operations explicit and improve readability.

4. Function Calls

Function-call syntax is function_name(argument1, argument2, ...). Calls may be nested. See Expression Operators for the complete function list.

Common function categories

Moving averages

ta_ma, ta_ema, ta_wma, t_mean

Momentum indicators

ta_rsi, ta_macd, ta_kdj, ta_cci

Volatility

ta_atr, ta_bbands, t_std

Time shifts

shift, pct, log, diff

Conditional functions

if, max, min, abs

Cross-sectional functions

c_rank, c_pctrank, c_zscore

Function-call examples

# Simple call
ta_ma(close, 20)// 20-period moving average

# Nested call
t_mean(ta_rsi(close, 14), 20)// 20-period average of RSI

# Conditional function
if(close > open, 1, -1)// Returns 1 for a bullish candle and -1 for a bearish candle

5. Factor Expression Examples

The following expressions are commonly used in the factor configuration of a data-processing node:

Factor	Expression	Description
Moving average	`ma20 = ta_ma(close, 20)`	20-period simple moving average
RSI	`rsi = ta_rsi(close, 14)`	14-period Relative Strength Index
Return	`ret = pct(close, 1)`	Single-period return
Volatility	`vol = t_std(pct(close, 1), 20)`	20-period standard deviation of returns
Volume ratio	`vol_ratio = volume / t_mean(volume, 20)`	Volume divided by 20-period average volume
Momentum	`mom = close / shift(close, 20) - 1`	20-period price rate of change
Bollinger position	`bb_pos = (close - ta_bbands_lower(close, 20, 2)) / (ta_bbands_upper(close, 20, 2) - ta_bbands_lower(close, 20, 2))`	Price position within the Bollinger Bands (0–1)

6. Condition Expression Examples

Condition	Strategy condition expression	Prerequisite in data processing
RSI overbought	`rsi > 70`	`rsi = ta_rsi(close, 14)`
RSI oversold	`rsi < 30`	`rsi = ta_rsi(close, 14)`
Golden cross signal	`golden == True`	`golden = ta_cross_over(ma5, ma20)`
Death cross signal	`death == True`	`death = ta_cross_under(ma5, ma20)`
Break above upper band	`close > bb_upper`	`bb_upper = ta_bbands_upper(close, 20, 2)`
Price and volume breakout	`(close > open) & (vol_ratio > 1.5)`	`vol_ratio = volume / t_mean(volume, 20)`
Multiple conditions	`(rsi < 30) & (close > ma20) & (vol_ratio > 1)`	`rsi / ma20 / vol_ratio are all precomputed`
AI prediction signal	`pred > 0.6`	`pred is emitted by the AI training node`

💡 Tip: A Boolean factor column such as golden or death may be referenced directly (golden) or compared explicitly (golden == True). Both forms have the same effect.

7. Label Expression Examples

The following expressions are commonly used for label definitions in a data-processing node. A label defines an AI model's prediction target and usually uses shift() to access future data.

Label	Expression	Task type
Future return	`label = shift(close, -5) / shift(open, -1) - 1`	Regression / binary classification
Future maximum gain	`label = shift(t_max(high, 5), -5) / shift(open, -1) - 1`	Regression
Cross-sectional ranking	`label = shift(close, -6) / shift(open, -1) - 1`	Ranking (converted by the engine)

💡 About shift():shift(col, n) shifts data backward when n > 0 or forward when n < 0. For example, shift(close, -5) means “the closing price five periods later.”

Trading engine node

Expression operators

Expression Rules and Syntax

1. Use Cases

Expressions are BeeQuant's core language for describing factor calculations, trading conditions, and label definitions. They are primarily used in the following places:

Location	Purpose	Example	Supported features
Data-processing node	Factor expression	`rsi = ta_rsi(close, 14)`	Operator functions + operators
	Label expression	`label = shift(close, -5) / shift(open, -1) - 1`	Operator functions + operators
	Filter expression	`volume > 1000000`	Comparison + logical operations
Strategy-type node	Timing entry/exit expression	`(rsi > 70) & (ma5 > ma20)`	Comparisons + logical operations only
Strategy-type node	Asset-selection filter expression	`volume > 1000000`	Comparisons + logical operations only

2. Basic Syntax

An expression consists of column names, constants, operators, and function calls.

2.1 Built-in Columns

The data-processing and extraction node retrieves the following columns from the exchange. You can reference them directly in its factor, label, and filter expressions:

Column	Description	Example
`open`	Opening price	`(close - open) / open`
`high`	Highest price	`high - low`
`low`	Lowest price	`close - low`
`close`	Closing price (most commonly used)	`ta_ma(close, 20)`

2.2 Constants

Type	Description	Examples
Integer	Used for window lengths, period parameters, and similar values	`14`, `20`, `60`
Floating-point number	Used for thresholds, ratios, and similar values	`0.05`, `2.0`, `-0.5`
Boolean	True/false	`True / 1`, `False / 0`

2.3 Variable Definitions

In factor configuration, define variables as variable_name = expression. Put each variable on a separate line.

Syntax

Example output from the configuration above (_ma20 is an intermediate variable and is omitted; diff and rsi are output):

timestamp	symbol	close	diff	rsi
2024-12-11 00:00:00	BTCUSDT	97800.0	-679.5	58.32
2024-12-11 01:00:00	BTCUSDT	97750.0	-593.2	55.18
2024-12-11 02:00:00	BTCUSDT	97820.0	-616.8	59.45
2024-12-11 03:00:00	BTCUSDT	98050.0	-781.5	63.21

● Gold columns are newly added output columns. _ma20 is omitted because its name begins with _.

💡 When to use intermediate variables:They make complex factor calculations easier to read.

3. Operators

3.1 Arithmetic Operators

Symbol	Name	Example	Description
+	Addition	`close + open`	Adds two values
-	Subtraction	`high - low`	Subtracts one value from another
*	Multiplication	`close * volume`	Multiplies two values
/	Division	`(close - open) / open`	Divides one value by another
**	Exponentiation	`returns ** 2`	Raises x to the power y
%	Modulo	`index % 5`	Returns the remainder after division

3.2 Comparison Operators

Comparison operators return Boolean values and are used primarily in strategy-node condition expressions.

Symbol	Name	Example	Description
>	Greater than	`rsi > 70`	RSI is overbought
>=	Greater than or equal to	`close >= ma20`	Price is at or above the moving average
<	Less than	`rsi < 30`	RSI is oversold
<=	Less than or equal to	`close <= ma20`	Price is at or below the moving average
==	Equal to	`signal == 1`	Signal equals 1
!=	Not equal to	`trend != 0`	Trend is not 0

3.3 Logical Operators

Logical operators combine multiple conditions.

Symbol	Name	Example	Description
&	AND	`(rsi < 30) & (close > ma20)`	Both conditions must be true
\|	OR	`(rsi > 70) \| (rsi < 30)`	Either condition may be true
~	NOT	`~(close > open)`	Negates the condition

⚠️ Note: When using & or |, wrap each condition in parentheses to avoid operator-precedence issues.

Operator precedence (highest to lowest)

1. ** Exponentiation

2. ~ Logical NOT

3. * / % Multiplication, division, modulo

4. + - Addition and subtraction

5. > >= < <= Comparisons

6. == != Equality tests

7. & Logical AND

8. | Logical OR

💡 Use parentheses to make the order of operations explicit and improve readability.

4. Function Calls

Function-call syntax is function_name(argument1, argument2, ...). Calls may be nested. See Expression Operators for the complete function list.

Common function categories

Moving averages

ta_ma, ta_ema, ta_wma, t_mean

Momentum indicators

ta_rsi, ta_macd, ta_kdj, ta_cci

Volatility

ta_atr, ta_bbands, t_std

Time shifts

shift, pct, log, diff

Conditional functions

if, max, min, abs

Cross-sectional functions

c_rank, c_pctrank, c_zscore

Function-call examples

# Simple call
ta_ma(close, 20)// 20-period moving average

# Nested call
t_mean(ta_rsi(close, 14), 20)// 20-period average of RSI

# Conditional function
if(close > open, 1, -1)// Returns 1 for a bullish candle and -1 for a bearish candle

5. Factor Expression Examples

The following expressions are commonly used in the factor configuration of a data-processing node:

Factor	Expression	Description
Moving average	`ma20 = ta_ma(close, 20)`	20-period simple moving average
RSI	`rsi = ta_rsi(close, 14)`	14-period Relative Strength Index
Return	`ret = pct(close, 1)`	Single-period return
Volatility	`vol = t_std(pct(close, 1), 20)`	20-period standard deviation of returns
Volume ratio	`vol_ratio = volume / t_mean(volume, 20)`	Volume divided by 20-period average volume
Momentum	`mom = close / shift(close, 20) - 1`	20-period price rate of change
Bollinger position	`bb_pos = (close - ta_bbands_lower(close, 20, 2)) / (ta_bbands_upper(close, 20, 2) - ta_bbands_lower(close, 20, 2))`	Price position within the Bollinger Bands (0–1)

6. Condition Expression Examples

Condition	Strategy condition expression	Prerequisite in data processing
RSI overbought	`rsi > 70`	`rsi = ta_rsi(close, 14)`
RSI oversold	`rsi < 30`	`rsi = ta_rsi(close, 14)`
Golden cross signal	`golden == True`	`golden = ta_cross_over(ma5, ma20)`
Death cross signal	`death == True`	`death = ta_cross_under(ma5, ma20)`
Break above upper band	`close > bb_upper`	`bb_upper = ta_bbands_upper(close, 20, 2)`
Price and volume breakout	`(close > open) & (vol_ratio > 1.5)`	`vol_ratio = volume / t_mean(volume, 20)`
Multiple conditions	`(rsi < 30) & (close > ma20) & (vol_ratio > 1)`	`rsi / ma20 / vol_ratio are all precomputed`
AI prediction signal	`pred > 0.6`	`pred is emitted by the AI training node`

💡 Tip: A Boolean factor column such as golden or death may be referenced directly (golden) or compared explicitly (golden == True). Both forms have the same effect.

7. Label Expression Examples

The following expressions are commonly used for label definitions in a data-processing node. A label defines an AI model's prediction target and usually uses shift() to access future data.

Label	Expression	Task type
Future return	`label = shift(close, -5) / shift(open, -1) - 1`	Regression / binary classification
Future maximum gain	`label = shift(t_max(high, 5), -5) / shift(open, -1) - 1`	Regression
Cross-sectional ranking	`label = shift(close, -6) / shift(open, -1) - 1`	Ranking (converted by the engine)

💡 About shift():shift(col, n) shifts data backward when n > 0 or forward when n < 0. For example, shift(close, -5) means “the closing price five periods later.”

Expression Rules and Syntax

1. Use Cases

Expressions are BeeQuant's core language for describing factor calculations, trading conditions, and label definitions. They are primarily used in the following places:

Location	Purpose	Example	Supported features
Data-processing node	Factor expression	`rsi = ta_rsi(close, 14)`	Operator functions + operators
	Label expression	`label = shift(close, -5) / shift(open, -1) - 1`	Operator functions + operators
	Filter expression	`volume > 1000000`	Comparison + logical operations
Strategy-type node	Timing entry/exit expression	`(rsi > 70) & (ma5 > ma20)`	Comparisons + logical operations only
Strategy-type node	Asset-selection filter expression	`volume > 1000000`	Comparisons + logical operations only

2. Basic Syntax

An expression consists of column names, constants, operators, and function calls.

2.1 Built-in Columns

The data-processing and extraction node retrieves the following columns from the exchange. You can reference them directly in its factor, label, and filter expressions:

Column	Description	Example
`open`	Opening price	`(close - open) / open`
`high`	Highest price	`high - low`
`low`	Lowest price	`close - low`
`close`	Closing price (most commonly used)	`ta_ma(close, 20)`

2.2 Constants

Type	Description	Examples
Integer	Used for window lengths, period parameters, and similar values	`14`, `20`, `60`
Floating-point number	Used for thresholds, ratios, and similar values	`0.05`, `2.0`, `-0.5`
Boolean	True/false	`True / 1`, `False / 0`

2.3 Variable Definitions

In factor configuration, define variables as variable_name = expression. Put each variable on a separate line.

Syntax

Example output from the configuration above (_ma20 is an intermediate variable and is omitted; diff and rsi are output):

timestamp	symbol	close	diff	rsi
2024-12-11 00:00:00	BTCUSDT	97800.0	-679.5	58.32
2024-12-11 01:00:00	BTCUSDT	97750.0	-593.2	55.18
2024-12-11 02:00:00	BTCUSDT	97820.0	-616.8	59.45
2024-12-11 03:00:00	BTCUSDT	98050.0	-781.5	63.21

● Gold columns are newly added output columns. _ma20 is omitted because its name begins with _.

💡 When to use intermediate variables:They make complex factor calculations easier to read.

3. Operators

3.1 Arithmetic Operators

Symbol	Name	Example	Description
+	Addition	`close + open`	Adds two values
-	Subtraction	`high - low`	Subtracts one value from another
*	Multiplication	`close * volume`	Multiplies two values
/	Division	`(close - open) / open`	Divides one value by another
**	Exponentiation	`returns ** 2`	Raises x to the power y
%	Modulo	`index % 5`	Returns the remainder after division

3.2 Comparison Operators

Comparison operators return Boolean values and are used primarily in strategy-node condition expressions.

Symbol	Name	Example	Description
>	Greater than	`rsi > 70`	RSI is overbought
>=	Greater than or equal to	`close >= ma20`	Price is at or above the moving average
<	Less than	`rsi < 30`	RSI is oversold
<=	Less than or equal to	`close <= ma20`	Price is at or below the moving average
==	Equal to	`signal == 1`	Signal equals 1
!=	Not equal to	`trend != 0`	Trend is not 0

3.3 Logical Operators

Logical operators combine multiple conditions.

Symbol	Name	Example	Description
&	AND	`(rsi < 30) & (close > ma20)`	Both conditions must be true
\|	OR	`(rsi > 70) \| (rsi < 30)`	Either condition may be true
~	NOT	`~(close > open)`	Negates the condition

⚠️ Note: When using & or |, wrap each condition in parentheses to avoid operator-precedence issues.

Operator precedence (highest to lowest)

1. ** Exponentiation

2. ~ Logical NOT

3. * / % Multiplication, division, modulo

4. + - Addition and subtraction

5. > >= < <= Comparisons

6. == != Equality tests

7. & Logical AND

8. | Logical OR

💡 Use parentheses to make the order of operations explicit and improve readability.

4. Function Calls

Function-call syntax is function_name(argument1, argument2, ...). Calls may be nested. See Expression Operators for the complete function list.

Common function categories

Moving averages

ta_ma, ta_ema, ta_wma, t_mean

Momentum indicators

ta_rsi, ta_macd, ta_kdj, ta_cci

Volatility

ta_atr, ta_bbands, t_std

Time shifts

shift, pct, log, diff

Conditional functions

if, max, min, abs

Cross-sectional functions

c_rank, c_pctrank, c_zscore

Function-call examples

# Simple call
ta_ma(close, 20)// 20-period moving average

# Nested call
t_mean(ta_rsi(close, 14), 20)// 20-period average of RSI

# Conditional function
if(close > open, 1, -1)// Returns 1 for a bullish candle and -1 for a bearish candle

5. Factor Expression Examples

The following expressions are commonly used in the factor configuration of a data-processing node:

Factor	Expression	Description
Moving average	`ma20 = ta_ma(close, 20)`	20-period simple moving average
RSI	`rsi = ta_rsi(close, 14)`	14-period Relative Strength Index
Return	`ret = pct(close, 1)`	Single-period return
Volatility	`vol = t_std(pct(close, 1), 20)`	20-period standard deviation of returns
Volume ratio	`vol_ratio = volume / t_mean(volume, 20)`	Volume divided by 20-period average volume
Momentum	`mom = close / shift(close, 20) - 1`	20-period price rate of change
Bollinger position	`bb_pos = (close - ta_bbands_lower(close, 20, 2)) / (ta_bbands_upper(close, 20, 2) - ta_bbands_lower(close, 20, 2))`	Price position within the Bollinger Bands (0–1)

6. Condition Expression Examples

Condition	Strategy condition expression	Prerequisite in data processing
RSI overbought	`rsi > 70`	`rsi = ta_rsi(close, 14)`
RSI oversold	`rsi < 30`	`rsi = ta_rsi(close, 14)`
Golden cross signal	`golden == True`	`golden = ta_cross_over(ma5, ma20)`
Death cross signal	`death == True`	`death = ta_cross_under(ma5, ma20)`
Break above upper band	`close > bb_upper`	`bb_upper = ta_bbands_upper(close, 20, 2)`
Price and volume breakout	`(close > open) & (vol_ratio > 1.5)`	`vol_ratio = volume / t_mean(volume, 20)`
Multiple conditions	`(rsi < 30) & (close > ma20) & (vol_ratio > 1)`	`rsi / ma20 / vol_ratio are all precomputed`
AI prediction signal	`pred > 0.6`	`pred is emitted by the AI training node`

💡 Tip: A Boolean factor column such as golden or death may be referenced directly (golden) or compared explicitly (golden == True). Both forms have the same effect.

7. Label Expression Examples

The following expressions are commonly used for label definitions in a data-processing node. A label defines an AI model's prediction target and usually uses shift() to access future data.

Label	Expression	Task type
Future return	`label = shift(close, -5) / shift(open, -1) - 1`	Regression / binary classification
Future maximum gain	`label = shift(t_max(high, 5), -5) / shift(open, -1) - 1`	Regression
Cross-sectional ranking	`label = shift(close, -6) / shift(open, -1) - 1`	Ranking (converted by the engine)

💡 About shift():shift(col, n) shifts data backward when n > 0 or forward when n < 0. For example, shift(close, -5) means “the closing price five periods later.”

Expression Rules and Syntax

1. Use Cases

Expressions are BeeQuant's core language for describing factor calculations, trading conditions, and label definitions. They are primarily used in the following places:

Location	Purpose	Example	Supported features
Data-processing node	Factor expression	`rsi = ta_rsi(close, 14)`	Operator functions + operators
	Label expression	`label = shift(close, -5) / shift(open, -1) - 1`	Operator functions + operators
	Filter expression	`volume > 1000000`	Comparison + logical operations
Strategy-type node	Timing entry/exit expression	`(rsi > 70) & (ma5 > ma20)`	Comparisons + logical operations only
Strategy-type node	Asset-selection filter expression	`volume > 1000000`	Comparisons + logical operations only

2. Basic Syntax

An expression consists of column names, constants, operators, and function calls.

2.1 Built-in Columns

The data-processing and extraction node retrieves the following columns from the exchange. You can reference them directly in its factor, label, and filter expressions:

Column	Description	Example
`open`	Opening price	`(close - open) / open`
`high`	Highest price	`high - low`
`low`	Lowest price	`close - low`
`close`	Closing price (most commonly used)	`ta_ma(close, 20)`

2.2 Constants

Type	Description	Examples
Integer	Used for window lengths, period parameters, and similar values	`14`, `20`, `60`
Floating-point number	Used for thresholds, ratios, and similar values	`0.05`, `2.0`, `-0.5`
Boolean	True/false	`True / 1`, `False / 0`

2.3 Variable Definitions

In factor configuration, define variables as variable_name = expression. Put each variable on a separate line.

Syntax

Example output from the configuration above (_ma20 is an intermediate variable and is omitted; diff and rsi are output):

timestamp	symbol	close	diff	rsi
2024-12-11 00:00:00	BTCUSDT	97800.0	-679.5	58.32
2024-12-11 01:00:00	BTCUSDT	97750.0	-593.2	55.18
2024-12-11 02:00:00	BTCUSDT	97820.0	-616.8	59.45
2024-12-11 03:00:00	BTCUSDT	98050.0	-781.5	63.21

● Gold columns are newly added output columns. _ma20 is omitted because its name begins with _.

💡 When to use intermediate variables:They make complex factor calculations easier to read.

3. Operators

3.1 Arithmetic Operators

Symbol	Name	Example	Description
+	Addition	`close + open`	Adds two values
-	Subtraction	`high - low`	Subtracts one value from another
*	Multiplication	`close * volume`	Multiplies two values
/	Division	`(close - open) / open`	Divides one value by another
**	Exponentiation	`returns ** 2`	Raises x to the power y
%	Modulo	`index % 5`	Returns the remainder after division

3.2 Comparison Operators

Comparison operators return Boolean values and are used primarily in strategy-node condition expressions.

Symbol	Name	Example	Description
>	Greater than	`rsi > 70`	RSI is overbought
>=	Greater than or equal to	`close >= ma20`	Price is at or above the moving average
<	Less than	`rsi < 30`	RSI is oversold
<=	Less than or equal to	`close <= ma20`	Price is at or below the moving average
==	Equal to	`signal == 1`	Signal equals 1
!=	Not equal to	`trend != 0`	Trend is not 0

3.3 Logical Operators

Logical operators combine multiple conditions.

Symbol	Name	Example	Description
&	AND	`(rsi < 30) & (close > ma20)`	Both conditions must be true
\|	OR	`(rsi > 70) \| (rsi < 30)`	Either condition may be true
~	NOT	`~(close > open)`	Negates the condition

⚠️ Note: When using & or |, wrap each condition in parentheses to avoid operator-precedence issues.

Operator precedence (highest to lowest)

1. ** Exponentiation

2. ~ Logical NOT

3. * / % Multiplication, division, modulo

4. + - Addition and subtraction

5. > >= < <= Comparisons

6. == != Equality tests

7. & Logical AND

8. | Logical OR

💡 Use parentheses to make the order of operations explicit and improve readability.

4. Function Calls

Function-call syntax is function_name(argument1, argument2, ...). Calls may be nested. See Expression Operators for the complete function list.

Common function categories

Moving averages

ta_ma, ta_ema, ta_wma, t_mean

Momentum indicators

ta_rsi, ta_macd, ta_kdj, ta_cci

Volatility

ta_atr, ta_bbands, t_std

Time shifts

shift, pct, log, diff

Conditional functions

if, max, min, abs

Cross-sectional functions

c_rank, c_pctrank, c_zscore

Function-call examples

# Simple call
ta_ma(close, 20)// 20-period moving average

# Nested call
t_mean(ta_rsi(close, 14), 20)// 20-period average of RSI

# Conditional function
if(close > open, 1, -1)// Returns 1 for a bullish candle and -1 for a bearish candle

5. Factor Expression Examples

The following expressions are commonly used in the factor configuration of a data-processing node:

Factor	Expression	Description
Moving average	`ma20 = ta_ma(close, 20)`	20-period simple moving average
RSI	`rsi = ta_rsi(close, 14)`	14-period Relative Strength Index
Return	`ret = pct(close, 1)`	Single-period return
Volatility	`vol = t_std(pct(close, 1), 20)`	20-period standard deviation of returns
Volume ratio	`vol_ratio = volume / t_mean(volume, 20)`	Volume divided by 20-period average volume
Momentum	`mom = close / shift(close, 20) - 1`	20-period price rate of change
Bollinger position	`bb_pos = (close - ta_bbands_lower(close, 20, 2)) / (ta_bbands_upper(close, 20, 2) - ta_bbands_lower(close, 20, 2))`	Price position within the Bollinger Bands (0–1)

6. Condition Expression Examples

Condition	Strategy condition expression	Prerequisite in data processing
RSI overbought	`rsi > 70`	`rsi = ta_rsi(close, 14)`
RSI oversold	`rsi < 30`	`rsi = ta_rsi(close, 14)`
Golden cross signal	`golden == True`	`golden = ta_cross_over(ma5, ma20)`
Death cross signal	`death == True`	`death = ta_cross_under(ma5, ma20)`
Break above upper band	`close > bb_upper`	`bb_upper = ta_bbands_upper(close, 20, 2)`
Price and volume breakout	`(close > open) & (vol_ratio > 1.5)`	`vol_ratio = volume / t_mean(volume, 20)`
Multiple conditions	`(rsi < 30) & (close > ma20) & (vol_ratio > 1)`	`rsi / ma20 / vol_ratio are all precomputed`
AI prediction signal	`pred > 0.6`	`pred is emitted by the AI training node`

💡 Tip: A Boolean factor column such as golden or death may be referenced directly (golden) or compared explicitly (golden == True). Both forms have the same effect.

7. Label Expression Examples

The following expressions are commonly used for label definitions in a data-processing node. A label defines an AI model's prediction target and usually uses shift() to access future data.

Label	Expression	Task type
Future return	`label = shift(close, -5) / shift(open, -1) - 1`	Regression / binary classification
Future maximum gain	`label = shift(t_max(high, 5), -5) / shift(open, -1) - 1`	Regression
Cross-sectional ranking	`label = shift(close, -6) / shift(open, -1) - 1`	Ranking (converted by the engine)

💡 About shift():shift(col, n) shifts data backward when n > 0 or forward when n < 0. For example, shift(close, -5) means “the closing price five periods later.”

Trading engine node

Expression operators

Expression Rules and Syntax

1. Use Cases

Expressions are BeeQuant's core language for describing factor calculations, trading conditions, and label definitions. They are primarily used in the following places:

Location	Purpose	Example	Supported features
Data-processing node	Factor expression	`rsi = ta_rsi(close, 14)`	Operator functions + operators
	Label expression	`label = shift(close, -5) / shift(open, -1) - 1`	Operator functions + operators
	Filter expression	`volume > 1000000`	Comparison + logical operations
Strategy-type node	Timing entry/exit expression	`(rsi > 70) & (ma5 > ma20)`	Comparisons + logical operations only
Strategy-type node	Asset-selection filter expression	`volume > 1000000`	Comparisons + logical operations only

2. Basic Syntax

An expression consists of column names, constants, operators, and function calls.

2.1 Built-in Columns

The data-processing and extraction node retrieves the following columns from the exchange. You can reference them directly in its factor, label, and filter expressions:

Column	Description	Example
`open`	Opening price	`(close - open) / open`
`high`	Highest price	`high - low`
`low`	Lowest price	`close - low`
`close`	Closing price (most commonly used)	`ta_ma(close, 20)`

2.2 Constants

Type	Description	Examples
Integer	Used for window lengths, period parameters, and similar values	`14`, `20`, `60`
Floating-point number	Used for thresholds, ratios, and similar values	`0.05`, `2.0`, `-0.5`
Boolean	True/false	`True / 1`, `False / 0`

2.3 Variable Definitions

In factor configuration, define variables as variable_name = expression. Put each variable on a separate line.

Syntax

Example output from the configuration above (_ma20 is an intermediate variable and is omitted; diff and rsi are output):

timestamp	symbol	close	diff	rsi
2024-12-11 00:00:00	BTCUSDT	97800.0	-679.5	58.32
2024-12-11 01:00:00	BTCUSDT	97750.0	-593.2	55.18
2024-12-11 02:00:00	BTCUSDT	97820.0	-616.8	59.45
2024-12-11 03:00:00	BTCUSDT	98050.0	-781.5	63.21

● Gold columns are newly added output columns. _ma20 is omitted because its name begins with _.

💡 When to use intermediate variables:They make complex factor calculations easier to read.

3. Operators

3.1 Arithmetic Operators

Symbol	Name	Example	Description
+	Addition	`close + open`	Adds two values
-	Subtraction	`high - low`	Subtracts one value from another
*	Multiplication	`close * volume`	Multiplies two values
/	Division	`(close - open) / open`	Divides one value by another
**	Exponentiation	`returns ** 2`	Raises x to the power y
%	Modulo	`index % 5`	Returns the remainder after division

3.2 Comparison Operators

Comparison operators return Boolean values and are used primarily in strategy-node condition expressions.

Symbol	Name	Example	Description
>	Greater than	`rsi > 70`	RSI is overbought
>=	Greater than or equal to	`close >= ma20`	Price is at or above the moving average
<	Less than	`rsi < 30`	RSI is oversold
<=	Less than or equal to	`close <= ma20`	Price is at or below the moving average
==	Equal to	`signal == 1`	Signal equals 1
!=	Not equal to	`trend != 0`	Trend is not 0

3.3 Logical Operators

Logical operators combine multiple conditions.

Symbol	Name	Example	Description
&	AND	`(rsi < 30) & (close > ma20)`	Both conditions must be true
\|	OR	`(rsi > 70) \| (rsi < 30)`	Either condition may be true
~	NOT	`~(close > open)`	Negates the condition

⚠️ Note: When using & or |, wrap each condition in parentheses to avoid operator-precedence issues.

Operator precedence (highest to lowest)

1. ** Exponentiation

2. ~ Logical NOT

3. * / % Multiplication, division, modulo

4. + - Addition and subtraction

5. > >= < <= Comparisons

6. == != Equality tests

7. & Logical AND

8. | Logical OR

💡 Use parentheses to make the order of operations explicit and improve readability.

4. Function Calls

Function-call syntax is function_name(argument1, argument2, ...). Calls may be nested. See Expression Operators for the complete function list.

Common function categories

Moving averages

ta_ma, ta_ema, ta_wma, t_mean

Momentum indicators

ta_rsi, ta_macd, ta_kdj, ta_cci

Volatility

ta_atr, ta_bbands, t_std

Time shifts

shift, pct, log, diff

Conditional functions

if, max, min, abs

Cross-sectional functions

c_rank, c_pctrank, c_zscore

Function-call examples

# Simple call
ta_ma(close, 20)// 20-period moving average

# Nested call
t_mean(ta_rsi(close, 14), 20)// 20-period average of RSI

# Conditional function
if(close > open, 1, -1)// Returns 1 for a bullish candle and -1 for a bearish candle

5. Factor Expression Examples

The following expressions are commonly used in the factor configuration of a data-processing node:

Factor	Expression	Description
Moving average	`ma20 = ta_ma(close, 20)`	20-period simple moving average
RSI	`rsi = ta_rsi(close, 14)`	14-period Relative Strength Index
Return	`ret = pct(close, 1)`	Single-period return
Volatility	`vol = t_std(pct(close, 1), 20)`	20-period standard deviation of returns
Volume ratio	`vol_ratio = volume / t_mean(volume, 20)`	Volume divided by 20-period average volume
Momentum	`mom = close / shift(close, 20) - 1`	20-period price rate of change
Bollinger position	`bb_pos = (close - ta_bbands_lower(close, 20, 2)) / (ta_bbands_upper(close, 20, 2) - ta_bbands_lower(close, 20, 2))`	Price position within the Bollinger Bands (0–1)

6. Condition Expression Examples

Condition	Strategy condition expression	Prerequisite in data processing
RSI overbought	`rsi > 70`	`rsi = ta_rsi(close, 14)`
RSI oversold	`rsi < 30`	`rsi = ta_rsi(close, 14)`
Golden cross signal	`golden == True`	`golden = ta_cross_over(ma5, ma20)`
Death cross signal	`death == True`	`death = ta_cross_under(ma5, ma20)`
Break above upper band	`close > bb_upper`	`bb_upper = ta_bbands_upper(close, 20, 2)`
Price and volume breakout	`(close > open) & (vol_ratio > 1.5)`	`vol_ratio = volume / t_mean(volume, 20)`
Multiple conditions	`(rsi < 30) & (close > ma20) & (vol_ratio > 1)`	`rsi / ma20 / vol_ratio are all precomputed`
AI prediction signal	`pred > 0.6`	`pred is emitted by the AI training node`

💡 Tip: A Boolean factor column such as golden or death may be referenced directly (golden) or compared explicitly (golden == True). Both forms have the same effect.

7. Label Expression Examples

The following expressions are commonly used for label definitions in a data-processing node. A label defines an AI model's prediction target and usually uses shift() to access future data.

Label	Expression	Task type
Future return	`label = shift(close, -5) / shift(open, -1) - 1`	Regression / binary classification
Future maximum gain	`label = shift(t_max(high, 5), -5) / shift(open, -1) - 1`	Regression
Cross-sectional ranking	`label = shift(close, -6) / shift(open, -1) - 1`	Ranking (converted by the engine)

💡 About shift():shift(col, n) shifts data backward when n > 0 or forward when n < 0. For example, shift(close, -5) means “the closing price five periods later.”