Features Engineering - Candle¶

Understanding candlestick patterns and their characteristics is crucial for traders and quantitative analysts. Candlestick features provide valuable insights into market sentiment, volatility, and momentum. In this notebook, we will explore key candle-based features using the quantreo package.

Tips: These indicators are short-term by nature, but their values can be smoothed to extract more meaningful insights—such as applying a Simple Moving Average (SMA) on the spread or similar techniques.

In [1]:

# Import the Features Engineering Package from Quantreo
import quantreo.features_engineering as fe

# To display the graphics
import matplotlib.pyplot as plt
plt.style.use("seaborn-v0_8")

# Import the Features Engineering Package from Quantreo import quantreo.features_engineering as fe # To display the graphics import matplotlib.pyplot as plt plt.style.use("seaborn-v0_8")

In [2]:

# Import a dataset to test the functions and create new ones easily
from quantreo.datasets import load_generated_ohlcv
df = load_generated_ohlcv()
df = df.loc["2016"]

# Show the data
df

# Import a dataset to test the functions and create new ones easily from quantreo.datasets import load_generated_ohlcv df = load_generated_ohlcv() df = df.loc["2016"] # Show the data df

Out[2]:

	open	high	low	close	volume
time
2016-01-04 00:00:00	104.944241	105.312073	104.929735	105.232289	576.805768
2016-01-04 04:00:00	105.233361	105.252139	105.047564	105.149357	485.696723
2016-01-04 08:00:00	105.159851	105.384745	105.141110	105.330306	403.969745
2016-01-04 12:00:00	105.330306	105.505799	104.894155	104.923404	1436.917324
2016-01-04 16:00:00	104.914147	105.023293	104.913252	105.014347	1177.672605
...	...	...	...	...	...
2016-12-30 04:00:00	103.632257	103.711884	103.495896	103.564574	563.932484
2016-12-30 08:00:00	103.564574	103.629321	103.555581	103.616731	697.707475
2016-12-30 12:00:00	103.615791	103.628165	103.496810	103.515847	1768.926665
2016-12-30 16:00:00	103.515847	103.692243	103.400370	103.422193	921.437054
2016-12-30 20:00:00	103.420285	103.429955	103.195921	103.226868	764.291608

1548 rows × 5 columns

Basic Candle Information¶

The candle_information function provides essential insights about a candle by calculating the following features:

• The Candle Way: This feature returns -1 if the candle is red (indicating a negative variation from the open price to the close price) or 1 if the candle is green (indicating a positive variation from the open price to the close price).

• The Filling: This feature computes the ratio between the absolute difference from the open to the close price and the total range of the candle (i.e., the difference between the high and low prices). $\frac{| \text{close} - \text{open} |}{| \text{high} - \text{low} |}$

• The Amplitude: This feature measures the relative price movement of a candle compared to its average. $\frac{| \text{close} - \text{open} |}{\left( \frac{\text{open} + \text{close}}{2} \right)}$

In [3]:

df["candle_way"], df["filling"], df["amplitude"] = fe.candle.candle_information(df=df, open_col="open", high_col="high",
                                                                               low_col="low", close_col="close")

df[["candle_way", "filling", "amplitude"]]

df["candle_way"], df["filling"], df["amplitude"] = fe.candle.candle_information(df=df, open_col="open", high_col="high", low_col="low", close_col="close") df[["candle_way", "filling", "amplitude"]]

Out[3]:

	candle_way	filling	amplitude
time
2016-01-04 00:00:00	1	0.753388	0.002741
2016-01-04 04:00:00	-1	0.410628	0.000799
2016-01-04 08:00:00	1	0.699634	0.001620
2016-01-04 12:00:00	-1	0.665260	0.003871
2016-01-04 16:00:00	1	0.910569	0.000955
...	...	...	...
2016-12-30 04:00:00	-1	0.313364	0.000653
2016-12-30 08:00:00	1	0.707317	0.000503
2016-12-30 12:00:00	-1	0.760870	0.000965
2016-12-30 16:00:00	-1	0.320872	0.000905
2016-12-30 20:00:00	-1	0.826446	0.001872

1548 rows × 3 columns

In [4]:

plt.figure(figsize=(15,6))
plt.plot(df["filling"])
plt.title("Filling Feature", size=15)
plt.show()

plt.figure(figsize=(15,6)) plt.plot(df["filling"]) plt.title("Filling Feature", size=15) plt.show()

In [5]:

plt.figure(figsize=(15,6))
plt.plot(df["amplitude"])
plt.title("Amplitude Feature", size=15)
plt.show()

plt.figure(figsize=(15,6)) plt.plot(df["amplitude"]) plt.title("Amplitude Feature", size=15) plt.show()

Spread Calculation¶

The compute_spread function provides a simple yet valuable insight into the difference between the highest and lowest price within a given period. This metric is useful for analyzing intra-bar market volatility.

$$ \text{spread} = \text{high} - \text{low} $$

The higher the spread value is, the more the market is volatile.

In [6]:

df["spread"] = fe.candle.compute_spread(df=df, high_col="high", low_col="low")

df["spread"]

df["spread"] = fe.candle.compute_spread(df=df, high_col="high", low_col="low") df["spread"]

Out[6]:

time
2016-01-04 00:00:00    0.382338
2016-01-04 04:00:00    0.204575
2016-01-04 08:00:00    0.243635
2016-01-04 12:00:00    0.611644
2016-01-04 16:00:00    0.110041
                         ...   
2016-12-30 04:00:00    0.215988
2016-12-30 08:00:00    0.073740
2016-12-30 12:00:00    0.131355
2016-12-30 16:00:00    0.291873
2016-12-30 20:00:00    0.234034
Name: spread, Length: 1548, dtype: float64

In [7]:

plt.figure(figsize=(15,6))
plt.plot(df["spread"])
plt.title("Spread Feature", size=15)
plt.show()

plt.figure(figsize=(15,6)) plt.plot(df["spread"]) plt.title("Spread Feature", size=15) plt.show()

Price Distribution¶

The price_distribution function calculates the percentage of closing prices that lie within a given relative range of their rolling high-low interval.

This feature is especially helpful to identify price compression (when many prices cluster in the center of the range) or volatility bursts (when price spreads out to the extremes).

It works by:

• Computing the min and max price in each rolling window.
• Evaluating the number of prices that fall between two dynamic thresholds:
  • start_percentage (e.g., 0.25) of the range
  • end_percentage (e.g., 0.75) of the range

In [8]:

# Compute the percentage of closing prices in different range zones using Quantreo
df["0_to_25"] = fe.candle.price_distribution(df, col="close", window_size=60, start_percentage=0.0, end_percentage=0.25)
df["25_to_75"] = fe.candle.price_distribution(df, col="close", window_size=60, start_percentage=0.25, end_percentage=0.75)
df["75_to_100"] = fe.candle.price_distribution(df, col="close", window_size=60, start_percentage=0.75, end_percentage=1.0)

# Prepare the stacked layers
bottom_25 = df["0_to_25"]
middle = df["25_to_75"]
top = df["75_to_100"]

# Create the stacked area plot
fig, ax = plt.subplots(figsize=(15, 6))

ax.fill_between(df.index, 0, bottom_25, color='blue', alpha=0.5, label='0–25%')
ax.fill_between(df.index, bottom_25, bottom_25 + middle, color='orange', alpha=0.5, label='25–75%')
ax.fill_between(df.index, bottom_25 + middle, bottom_25 + middle + top, color='green', alpha=0.5, label='75–100%')

# Customize the plot
ax.set_ylim(0, 100)
ax.set_title("Close Price Distribution within Range Zones", fontsize=14)
ax.set_xlabel("Date")
ax.set_ylabel("Percentage of Closes in Zone")
ax.legend()
ax.grid(True)

plt.tight_layout()
plt.show()

# Compute the percentage of closing prices in different range zones using Quantreo df["0_to_25"] = fe.candle.price_distribution(df, col="close", window_size=60, start_percentage=0.0, end_percentage=0.25) df["25_to_75"] = fe.candle.price_distribution(df, col="close", window_size=60, start_percentage=0.25, end_percentage=0.75) df["75_to_100"] = fe.candle.price_distribution(df, col="close", window_size=60, start_percentage=0.75, end_percentage=1.0) # Prepare the stacked layers bottom_25 = df["0_to_25"] middle = df["25_to_75"] top = df["75_to_100"] # Create the stacked area plot fig, ax = plt.subplots(figsize=(15, 6)) ax.fill_between(df.index, 0, bottom_25, color='blue', alpha=0.5, label='0–25%') ax.fill_between(df.index, bottom_25, bottom_25 + middle, color='orange', alpha=0.5, label='25–75%') ax.fill_between(df.index, bottom_25 + middle, bottom_25 + middle + top, color='green', alpha=0.5, label='75–100%') # Customize the plot ax.set_ylim(0, 100) ax.set_title("Close Price Distribution within Range Zones", fontsize=14) ax.set_xlabel("Date") ax.set_ylabel("Percentage of Closes in Zone") ax.legend() ax.grid(True) plt.tight_layout() plt.show()

Internal Bar Strength¶

The internal_bar_strength function computes the relative position of the closing price within the bar's range, making it a useful feature for detecting short-term overbought or oversold conditions.

IBS is especially relevant in mean-reversion strategies or when identifying compression zones within a bar.

The formula used is:

$$ \text{IBS} = \frac{\text{Close} - \text{Low}}{\text{High} - \text{Low}} $$

• IBS close to 1 → the price closed near the high of the day
• IBS close to 0 → the price closed near the low of the day
• IBS around 0.5 → the price closed in the middle of the range

In [9]:

df["ibs"] = fe.candle.internal_bar_strength(df=df, high_col="high", low_col="low", close_col="close")
df["ibs"]

df["ibs"] = fe.candle.internal_bar_strength(df=df, high_col="high", low_col="low", close_col="close") df["ibs"]

Out[9]:

time
2016-01-04 00:00:00    0.791328
2016-01-04 04:00:00    0.497585
2016-01-04 08:00:00    0.776557
2016-01-04 12:00:00    0.047820
2016-01-04 16:00:00    0.918699
                         ...   
2016-12-30 04:00:00    0.317972
2016-12-30 08:00:00    0.829268
2016-12-30 12:00:00    0.144928
2016-12-30 16:00:00    0.074766
2016-12-30 20:00:00    0.132231
Name: ibs, Length: 1548, dtype: float64

In [10]:

plt.figure(figsize=(15,6))
plt.plot(df["ibs"])
plt.title("IBS Feature", size=15)
plt.show()

plt.figure(figsize=(15,6)) plt.plot(df["ibs"]) plt.title("IBS Feature", size=15) plt.show()