## Wednesday 27 November 2013

### Brownian Bands revisited?

Further to my most recent post there have been a few anonymous comments suggesting that Brownian bands are more useful than I supposed. In response I have added a contact form to this blog ( on the right ) so these anonymous readers may contact me if they wish. I am curious to learn how they may be using Brownian bands. I might also add that I have thought of creating a new indicator ( a kind of randomness measurement oscillator ) based on the Brownian bands concept.

## Sunday 24 November 2013

### Suspending Work on Brownian Bands

Recently I have been doing a lot of work looking at various uses of the Brownian bands but I have been unable to come up with anything unique that I feel can help me so for now I am going to discontinue this work. There may be a time in the future when I return to this, but for now the Brownian bands are going to be put on the shelf along with my other indicators.

I have also started to become disillusioned with my attempts at creating a market classifier neural net; I think it may very well be too large a project to complete satisfactorily, so I am going to try something simpler in concept at least. In large part this change of heart is due to some recent reading I have been doing, particularly this article in which I was struck by the phrase

Also over on the Mechanical Forex site there is a whole series of articles on neural nets which I have found useful and have given me a new insight. I am now going to ask readers to suspend their disbelief for a few moments while I explain

Imagine it is a few minutes before the trading open and you have to decide whether to enter long, short or remain out of the market when the market opens. However, unlike any other market participant, the Gods have given you a gift - the ability to see the near future - which means you know the OHLC for today, tomorrow and the next trading day. But, as with all gifts from the Gods, there is a catch - you can only enter and exit trades at the market open; there is no buying the low and selling the high of a candlestick bar or intraday trading permitted by the Gods. What would such a blessed but responsible trader do? How would they make their trade decision? Being responsible s/he might consider the reward to risk ratio and only take a trade if this ratio is greater than one: on the long side the maximum open of tomorrow or the day after minus today's open divided by today's open minus the minimum low of today, tomorrow or the day after: and a similar reasoning for the short side. If neither of these ratios is greater than one, no new trade will be entered today.

The upper pane in the video below shows the coding of such logic. The decision to go long is rendered in blue, short in red and neutral in green. The colour of the bar indicates the action to be taken at the open of the next bar. However, it might be that when a neutral signal is given there is already be a position held, in which case the existing position is held for the duration of the neutral signal. This is effectively an always in the market, stop and reverse signal, and this is shown in the lower pane of the video. To make things slightly easier to see the entry/exit action occurs at the open of a new colour bar, i.e. if the bars change from blue to red the long is exited and the short initiated at the open of the first red bar.

I have also started to become disillusioned with my attempts at creating a market classifier neural net; I think it may very well be too large a project to complete satisfactorily, so I am going to try something simpler in concept at least. In large part this change of heart is due to some recent reading I have been doing, particularly this article in which I was struck by the phrase

*"As an aside, the ideal "trades" formed by the turning points might make a good training set of trades for a neural network-based system. Rather than scanning a chart manually to come up with trades to feed into the neural network, the method described here could be used to automatically find the training set."*Also over on the Mechanical Forex site there is a whole series of articles on neural nets which I have found useful and have given me a new insight. I am now going to ask readers to suspend their disbelief for a few moments while I explain

*.*Imagine it is a few minutes before the trading open and you have to decide whether to enter long, short or remain out of the market when the market opens. However, unlike any other market participant, the Gods have given you a gift - the ability to see the near future - which means you know the OHLC for today, tomorrow and the next trading day. But, as with all gifts from the Gods, there is a catch - you can only enter and exit trades at the market open; there is no buying the low and selling the high of a candlestick bar or intraday trading permitted by the Gods. What would such a blessed but responsible trader do? How would they make their trade decision? Being responsible s/he might consider the reward to risk ratio and only take a trade if this ratio is greater than one: on the long side the maximum open of tomorrow or the day after minus today's open divided by today's open minus the minimum low of today, tomorrow or the day after: and a similar reasoning for the short side. If neither of these ratios is greater than one, no new trade will be entered today.

The upper pane in the video below shows the coding of such logic. The decision to go long is rendered in blue, short in red and neutral in green. The colour of the bar indicates the action to be taken at the open of the next bar. However, it might be that when a neutral signal is given there is already be a position held, in which case the existing position is held for the duration of the neutral signal. This is effectively an always in the market, stop and reverse signal, and this is shown in the lower pane of the video. To make things slightly easier to see the entry/exit action occurs at the open of a new colour bar, i.e. if the bars change from blue to red the long is exited and the short initiated at the open of the first red bar.

Even with a cursory viewing it can be seen what a great "system" this would be, and using neural nets to create such a "system" is now my ambition.

The plan is simple: roll a moving window along the price series and use the known relationships between bars and indicators within this window to train a locally optimised neural net. The purpose of the training will be to classify the bars as long entry, short entry or neutral as in the chart in the upper pane of the above video. At the hard right edge of the chart the last three bars will be unavailable to the neural net for training purposes, but the hope is that the neural net, sufficiently well trained on all data in the window immediately prior to these three bars, will have predictive ability for them. After all, in the main, market dynamics slowly evolve over a few bars rather than dramatically leap.

Before I embark on this new work there are a few optimisation tests I would like to conduct, and these tests will form the subject matter of my next few posts.

## Sunday 17 November 2013

### Brownian Bands

For the past few weeks I have been working on the idea presented in my earlier modelling prices using brownian motion post and below is a video of what I have come up with, which I'm calling Brownian Bands. The video shows these bands on the EURUSD forex pair with look back lengths of 1 to 10 bars inclusive, then 15 to 50 in increments of 5, then 50, 100 and 200 and finally 50, 100 and 200 on the whole time series. Each look back length band is the average of all bands from 1 up to and including the look back length band - the details of the implementation can be seen in the code boxes below. The blue band is the upper band, the red is the lower and the green is the mid point of the two.

Non-embedded view.

This following chart shows an adaptive look back length implementation, the look back length being determined by a period vector input to the function, which in this case is the dominant cycle period.

These adaptive Brownian bands can be compared with the equivalent adaptive Bollinger bands shown in this next chart.

For my intended purpose of separating out returns into distribution bins the Brownian bands seem to be superior. There are 3 distinct bins: above the upper band, below the lower band and between the bands, whereas with the Bollinger bands the vast majority of returns would be lumped into just one bin - between the bands. Of course the Brownian bands could be used to create a system indicator just as Bollinger bands can be, and if any readers test this idea I would be intrigued to hear of your results. However, should you choose to do this there is a major caveat you should be aware of. The calculation of the Brownian bands is based on the natural log of actual price, which means you should be very wary of test results on any back adjusted price series such as futures and adjusted stock prices. In fact I would go as far to say that the code given below should not be used as is on such price series. It is for this reason that I have been using forex price series in my own tests so far, but I will surely get around to adjusting the given code in due course.

Now for the code. This first code box shows a hard-coded, unrolled loop version for look back lengths up to and including 25 for the benefit of a reader who requested a clearer exposition of the method than was given in my earlier post, which was vectorised Octave code. All the code below is C++ as used in Octave .oct files.

#include octave/oct.h

#include octave/dColVector.h

#include math.h

with the customary < before "octave" and "math" and > after ".h"

Non-embedded view.

This following chart shows an adaptive look back length implementation, the look back length being determined by a period vector input to the function, which in this case is the dominant cycle period.

These adaptive Brownian bands can be compared with the equivalent adaptive Bollinger bands shown in this next chart.

For my intended purpose of separating out returns into distribution bins the Brownian bands seem to be superior. There are 3 distinct bins: above the upper band, below the lower band and between the bands, whereas with the Bollinger bands the vast majority of returns would be lumped into just one bin - between the bands. Of course the Brownian bands could be used to create a system indicator just as Bollinger bands can be, and if any readers test this idea I would be intrigued to hear of your results. However, should you choose to do this there is a major caveat you should be aware of. The calculation of the Brownian bands is based on the natural log of actual price, which means you should be very wary of test results on any back adjusted price series such as futures and adjusted stock prices. In fact I would go as far to say that the code given below should not be used as is on such price series. It is for this reason that I have been using forex price series in my own tests so far, but I will surely get around to adjusting the given code in due course.

Now for the code. This first code box shows a hard-coded, unrolled loop version for look back lengths up to and including 25 for the benefit of a reader who requested a clearer exposition of the method than was given in my earlier post, which was vectorised Octave code. All the code below is C++ as used in Octave .oct files.

```
DEFUN_DLD ( brownian_bands_25, args, nargout,
"-*- texinfo -*-\n\
@deftypefn {Function File} {} brownian_bands_25 (@var{price,period})\n\
This function takes a price vector input and a value for the tick\n\
size of the instrument and outputs upper and lower bands based on\n\
the concept of Brownian motion. The bands are the average of look\n\
back periods of 1 to 25 bars inclusive of the relevant calculations.\n\
@end deftypefn" )
{
octave_value_list retval_list ;
int nargin = args.length () ;
// check the input arguments
if ( nargin != 2 )
{
error ( "Invalid arguments. Type help to see usage." ) ;
return retval_list ;
}
if ( args(0).length () < 21 )
{
error ( "Invalid arguments. Type help to see usage." ) ;
return retval_list ;
}
if ( args(1).length () != 1 )
{
error ( "Invalid arguments. Type help to see usage." ) ;
return retval_list ;
}
if ( error_state )
{
error ( "Invalid arguments. Type help to see usage." ) ;
return retval_list ;
}
// end of input checking
ColumnVector price = args(0).column_vector_value () ;
double tick_size = args(1).double_value() ;
ColumnVector abs_log_price_diff = args(0).column_vector_value () ;
ColumnVector upper_band = args(0).column_vector_value () ;
ColumnVector lower_band = args(0).column_vector_value () ;
ColumnVector mid_band = args(0).column_vector_value () ;
double sum ;
double up_1 ;
double low_1 ;
double up_2 ;
double low_2 ;
double sqrt_2 = sqrt(2.0) ; // pre-calculated for speed optimisation
double up_3 ;
double low_3 ;
double sqrt_3 = sqrt(3.0) ; // pre-calculated for speed optimisation
double up_4 ;
double low_4 ;
double sqrt_4 = sqrt(4.0) ; // pre-calculated for speed optimisation
double up_5 ;
double low_5 ;
double sqrt_5 = sqrt(5.0) ; // pre-calculated for speed optimisation
double up_6 ;
double low_6 ;
double sqrt_6 = sqrt(6.0) ; // pre-calculated for speed optimisation
double up_7 ;
double low_7 ;
double sqrt_7 = sqrt(7.0) ; // pre-calculated for speed optimisation
double up_8 ;
double low_8 ;
double sqrt_8 = sqrt(8.0) ; // pre-calculated for speed optimisation
double up_9 ;
double low_9 ;
double sqrt_9 = sqrt(9.0) ; // pre-calculated for speed optimisation
double up_10 ;
double low_10 ;
double sqrt_10 = sqrt(10.0) ; // pre-calculated for speed optimisation
double up_11 ;
double low_11 ;
double sqrt_11 = sqrt(11.0) ; // pre-calculated for speed optimisation
double up_12 ;
double low_12 ;
double sqrt_12 = sqrt(12.0) ; // pre-calculated for speed optimisation
double up_13 ;
double low_13 ;
double sqrt_13 = sqrt(13.0) ; // pre-calculated for speed optimisation
double up_14 ;
double low_14 ;
double sqrt_14 = sqrt(14.0) ; // pre-calculated for speed optimisation
double up_15 ;
double low_15 ;
double sqrt_15 = sqrt(15.0) ; // pre-calculated for speed optimisation
double up_16 ;
double low_16 ;
double sqrt_16 = sqrt(16.0) ; // pre-calculated for speed optimisation
double up_17 ;
double low_17 ;
double sqrt_17 = sqrt(17.0) ; // pre-calculated for speed optimisation
double up_18 ;
double low_18 ;
double sqrt_18 = sqrt(18.0) ; // pre-calculated for speed optimisation
double up_19 ;
double low_19 ;
double sqrt_19 = sqrt(19.0) ; // pre-calculated for speed optimisation
double up_20 ;
double low_20 ;
double sqrt_20 = sqrt(20.0) ; // pre-calculated for speed optimisation
double up_21 ;
double low_21 ;
double sqrt_21 = sqrt(21.0) ; // pre-calculated for speed optimisation
double up_22 ;
double low_22 ;
double sqrt_22 = sqrt(22.0) ; // pre-calculated for speed optimisation
double up_23 ;
double low_23 ;
double sqrt_23 = sqrt(23.0) ; // pre-calculated for speed optimisation
double up_24 ;
double low_24 ;
double sqrt_24 = sqrt(24.0) ; // pre-calculated for speed optimisation
double up_25 ;
double low_25 ;
double sqrt_25 = sqrt(25.0) ; // pre-calculated for speed optimisation
for ( octave_idx_type ii (1) ; ii < 25 ; ii++ ) // initialising loop
{
abs_log_price_diff(ii) = fabs( log( price(ii) ) - log( price(ii-1) ) ) ;
}
for ( octave_idx_type ii (25) ; ii < args(0).length () ; ii++ ) // main loop
{
abs_log_price_diff(ii) = fabs( log( price(ii) ) - log( price(ii-1) ) ) ;
sum = abs_log_price_diff(ii) ;
up_1 = exp( log( price(ii-1) ) + sum ) ;
low_1 = exp( log( price(ii-1) ) - sum ) ;
sum += abs_log_price_diff(ii-1) ;
up_2 = exp( log( price(ii-2) ) + (sum/2.0) * sqrt_2 ) ;
low_2 = exp( log( price(ii-2) ) - (sum/2.0) * sqrt_2 ) ;
sum += abs_log_price_diff(ii-2) ;
up_3 = exp( log( price(ii-3) ) + (sum/3.0) * sqrt_3 ) ;
low_3 = exp( log( price(ii-3) ) - (sum/3.0) * sqrt_3 ) ;
sum += abs_log_price_diff(ii-3) ;
up_4 = exp( log( price(ii-4) ) + (sum/4.0) * sqrt_4 ) ;
low_4 = exp( log( price(ii-4) ) - (sum/4.0) * sqrt_4 ) ;
sum += abs_log_price_diff(ii-4) ;
up_5 = exp( log( price(ii-5) ) + (sum/5.0) * sqrt_5 ) ;
low_5 = exp( log( price(ii-5) ) - (sum/5.0) * sqrt_5 ) ;
sum += abs_log_price_diff(ii-5) ;
up_6 = exp( log( price(ii-6) ) + (sum/6.0) * sqrt_6 ) ;
low_6 = exp( log( price(ii-6) ) - (sum/6.0) * sqrt_6 ) ;
sum += abs_log_price_diff(ii-6) ;
up_7 = exp( log( price(ii-7) ) + (sum/7.0) * sqrt_7 ) ;
low_7 = exp( log( price(ii-7) ) - (sum/7.0) * sqrt_7 ) ;
sum += abs_log_price_diff(ii-7) ;
up_8 = exp( log( price(ii-8) ) + (sum/8.0) * sqrt_8 ) ;
low_8 = exp( log( price(ii-8) ) - (sum/8.0) * sqrt_8 ) ;
sum += abs_log_price_diff(ii-8) ;
up_9 = exp( log( price(ii-9) ) + (sum/9.0) * sqrt_9 ) ;
low_9 = exp( log( price(ii-9) ) - (sum/9.0) * sqrt_9 ) ;
sum += abs_log_price_diff(ii-9) ;
up_10 = exp( log( price(ii-10) ) + (sum/10.0) * sqrt_10 ) ;
low_10 = exp( log( price(ii-10) ) - (sum/10.0) * sqrt_10 ) ;
sum += abs_log_price_diff(ii-10) ;
up_11 = exp( log( price(ii-11) ) + (sum/11.0) * sqrt_11 ) ;
low_11 = exp( log( price(ii-11) ) - (sum/11.0) * sqrt_11 ) ;
sum += abs_log_price_diff(ii-11) ;
up_12 = exp( log( price(ii-12) ) + (sum/12.0) * sqrt_12 ) ;
low_12 = exp( log( price(ii-12) ) - (sum/12.0) * sqrt_12 ) ;
sum += abs_log_price_diff(ii-12) ;
up_13 = exp( log( price(ii-13) ) + (sum/13.0) * sqrt_13 ) ;
low_13 = exp( log( price(ii-13) ) - (sum/13.0) * sqrt_13 ) ;
sum += abs_log_price_diff(ii-13) ;
up_14 = exp( log( price(ii-14) ) + (sum/14.0) * sqrt_14 ) ;
low_14 = exp( log( price(ii-14) ) - (sum/14.0) * sqrt_14 ) ;
sum += abs_log_price_diff(ii-14) ;
up_15 = exp( log( price(ii-15) ) + (sum/15.0) * sqrt_15 ) ;
low_15 = exp( log( price(ii-15) ) - (sum/15.0) * sqrt_15 ) ;
sum += abs_log_price_diff(ii-15) ;
up_16 = exp( log( price(ii-16) ) + (sum/16.0) * sqrt_16 ) ;
low_16 = exp( log( price(ii-16) ) - (sum/16.0) * sqrt_16 ) ;
sum += abs_log_price_diff(ii-16) ;
up_17 = exp( log( price(ii-17) ) + (sum/17.0) * sqrt_17 ) ;
low_17 = exp( log( price(ii-17) ) - (sum/17.0) * sqrt_17 ) ;
sum += abs_log_price_diff(ii-17) ;
up_18 = exp( log( price(ii-18) ) + (sum/18.0) * sqrt_18 ) ;
low_18 = exp( log( price(ii-18) ) - (sum/18.0) * sqrt_18 ) ;
sum += abs_log_price_diff(ii-18) ;
up_19 = exp( log( price(ii-19) ) + (sum/19.0) * sqrt_19 ) ;
low_19 = exp( log( price(ii-19) ) - (sum/19.0) * sqrt_19 ) ;
sum += abs_log_price_diff(ii-19) ;
up_20 = exp( log( price(ii-20) ) + (sum/20.0) * sqrt_20 ) ;
low_20 = exp( log( price(ii-20) ) - (sum/20.0) * sqrt_20 ) ;
sum += abs_log_price_diff(ii-20) ;
up_21 = exp( log( price(ii-21) ) + (sum/21.0) * sqrt_21 ) ;
low_21 = exp( log( price(ii-21) ) - (sum/21.0) * sqrt_21 ) ;
sum += abs_log_price_diff(ii-21) ;
up_22 = exp( log( price(ii-22) ) + (sum/22.0) * sqrt_22 ) ;
low_22 = exp( log( price(ii-22) ) - (sum/22.0) * sqrt_22 ) ;
sum += abs_log_price_diff(ii-22) ;
up_23 = exp( log( price(ii-23) ) + (sum/23.0) * sqrt_23 ) ;
low_23 = exp( log( price(ii-23) ) - (sum/23.0) * sqrt_23 ) ;
sum += abs_log_price_diff(ii-23) ;
up_24 = exp( log( price(ii-24) ) + (sum/24.0) * sqrt_24 ) ;
low_24 = exp( log( price(ii-24) ) - (sum/24.0) * sqrt_24 ) ;
sum += abs_log_price_diff(ii-24) ;
up_25 = exp( log( price(ii-25) ) + (sum/25.0) * sqrt_25 ) ;
low_25 = exp( log( price(ii-25) ) - (sum/25.0) * sqrt_25 ) ;
upper_band(ii) = (up_1+up_2+up_3+up_4+up_5+up_6+up_7+up_8+up_9+up_10+up_11+up_12+up_13+up_14+up_15+up_16+up_17+up_18+up_19+up_20+up_21+up_22+up_23+up_24+up_25)/25.0 ;
lower_band(ii) = (low_1+low_2+low_3+low_4+low_5+low_6+low_7+low_8+low_9+low_10+low_11+low_12+low_13+low_14+low_15+low_16+low_17+low_18+low_19+low_20+low_21+low_22+low_23+up_24+low_25)/25.0 ;
// round the upper_band up to the nearest tick
upper_band(ii) = ceil( upper_band(ii)/tick_size ) * tick_size ;
// round the lower_band down to the nearest tick
lower_band(ii) = floor( lower_band(ii)/tick_size ) * tick_size ;
mid_band(ii) = ( upper_band(ii) + lower_band(ii) ) / 2.0 ;
} // end of main loop
retval_list(2) = mid_band ;
retval_list(1) = lower_band ;
retval_list(0) = upper_band ;
return retval_list ;
} // end of function
```

This next code box contains code for an adjustable look back length, the length being a user input to the function. This code wraps the above unrolled loop into a nested loop.
```
DEFUN_DLD ( brownian_bands_adjustable, args, nargout,
"-*- texinfo -*-\n\
@deftypefn {Function File} {} brownian_bands_adjustable (@var{price,lookback,tick_size})\n\
This function takes a price vector input, a value for the lookback\n\
length and a value for the tick size of the instrument and outputs\n\
upper and lower bands based on the concept of Brownian motion. The bands\n\
are the average of lookback period bands from 1 to lookback length.\n\
@end deftypefn" )
{
octave_value_list retval_list ;
int nargin = args.length () ;
// check the input arguments
if ( nargin != 3 )
{
error ( "Invalid arguments. Type help to see usage." ) ;
return retval_list ;
}
if ( args(1).length () != 1 ) // lookback length argument
{
error ( "Invalid arguments. Type help to see usage." ) ;
return retval_list ;
}
double lookback = args(1).double_value() ;
if ( args(0).length () < lookback ) // check length of price vector input
{
error ( "Invalid arguments. Type help to see usage." ) ;
return retval_list ;
}
if ( args(2).length () != 1 ) // tick size
{
error ( "Invalid arguments. Type help to see usage." ) ;
return retval_list ;
}
if ( error_state )
{
error ( "Invalid arguments. Type help to see usage." ) ;
return retval_list ;
}
// end of input checking
ColumnVector price = args(0).column_vector_value () ;
double tick_size = args(2).double_value() ;
ColumnVector abs_log_price_diff = args(0).column_vector_value () ;
ColumnVector upper_band = args(0).column_vector_value () ;
ColumnVector lower_band = args(0).column_vector_value () ;
ColumnVector mid_band = args(0).column_vector_value () ;
double sum ;
double up_bb ;
double low_bb ;
int jj ;
for ( octave_idx_type ii (1) ; ii < lookback ; ii++ ) // initialising loop
{
abs_log_price_diff(ii) = fabs( log( price(ii) ) - log( price(ii-1) ) ) ;
}
for ( octave_idx_type ii (lookback) ; ii < args(0).length () ; ii++ ) // main loop
{
// initialise calculation values
sum = 0.0 ;
up_bb = 0.0 ;
low_bb = 0.0 ;
for ( jj = 1 ; jj < lookback + 1 ; jj++ ) // nested jj loop
{
abs_log_price_diff(ii) = fabs( log( price(ii) ) - log( price(ii-1) ) ) ;
sum += abs_log_price_diff(ii-jj+1) ;
up_bb += exp( log( price(ii-jj) ) + (sum/double(jj)) * sqrt(double(jj)) ) ;
low_bb += exp( log( price(ii-jj) ) - (sum/double(jj)) * sqrt(double(jj)) ) ;
} // end of nested jj loop
upper_band(ii) = up_bb / lookback ;
lower_band(ii) = low_bb / lookback ;
// round the upper_band up to the nearest tick
upper_band(ii) = ceil( upper_band(ii)/tick_size ) * tick_size ;
// round the lower_band down to the nearest tick
lower_band(ii) = floor( lower_band(ii)/tick_size ) * tick_size ;
mid_band(ii) = ( upper_band(ii) + lower_band(ii) ) / 2.0 ;
} // end of main loop
retval_list(2) = mid_band ;
retval_list(1) = lower_band ;
retval_list(0) = upper_band ;
return retval_list ;
} // end of function
```

Finally, this code is the adaptive look back version where the length of the look back is contained in a vector which is also a user input.
```
DEFUN_DLD ( brownian_bands_adaptive, args, nargout,
"-*- texinfo -*-\n\
@deftypefn {Function File} {} brownian_bands_adaptive (@var{price,period,tick_size})\n\
This function takes price and period vector inputs & a value for the tick\n\
size of the instrument and outputs upper and lower bands based on\n\
the concept of Brownian motion. The bands are adaptive averages of look\n\
back lengths of 1 to period bars inclusive.\n\
@end deftypefn" )
{
octave_value_list retval_list ;
int nargin = args.length () ;
// check the input arguments
if ( nargin != 3 )
{
error ( "Invalid arguments. Type help to see usage." ) ;
return retval_list ;
}
if ( args(0).length () < 50 ) // check length of price vector input
{
error ( "Invalid arguments. Type help to see usage." ) ;
return retval_list ;
}
if ( args(1).length () != args(0).length () ) // lookback period length argument
{
error ( "Invalid arguments. Type help to see usage." ) ;
return retval_list ;
}
if ( args(2).length () != 1 ) // tick size
{
error ( "Invalid arguments. Type help to see usage." ) ;
return retval_list ;
}
if ( error_state )
{
error ( "Invalid arguments. Type help to see usage." ) ;
return retval_list ;
}
// end of input checking
ColumnVector price = args(0).column_vector_value () ;
ColumnVector period = args(1).column_vector_value () ;
double tick_size = args(2).double_value() ;
ColumnVector abs_log_price_diff = args(0).column_vector_value () ;
ColumnVector upper_band = args(0).column_vector_value () ;
ColumnVector lower_band = args(0).column_vector_value () ;
ColumnVector mid_band = args(0).column_vector_value () ;
double sum ;
double up_bb ;
double low_bb ;
int jj ;
for ( octave_idx_type ii (1) ; ii < 50 ; ii++ ) // initialising loop
{
abs_log_price_diff(ii) = fabs( log( price(ii) ) - log( price(ii-1) ) ) ;
}
for ( octave_idx_type ii (50) ; ii < args(0).length () ; ii++ ) // main loop
{
// initialise calculation values
sum = 0.0 ;
up_bb = 0.0 ;
low_bb = 0.0 ;
for ( jj = 1 ; jj < period(ii) + 1 ; jj++ ) // nested jj loop
{
abs_log_price_diff(ii) = fabs( log( price(ii) ) - log( price(ii-1) ) ) ;
sum += abs_log_price_diff(ii-jj+1) ;
up_bb += exp( log( price(ii-jj) ) + (sum/double(jj)) * sqrt(double(jj)) ) ;
low_bb += exp( log( price(ii-jj) ) - (sum/double(jj)) * sqrt(double(jj)) ) ;
} // end of nested jj loop
upper_band(ii) = up_bb / period(ii) ;
lower_band(ii) = low_bb / period(ii) ;
// round the upper_band up to the nearest tick
upper_band(ii) = ceil( upper_band(ii)/tick_size ) * tick_size ;
// round the lower_band down to the nearest tick
lower_band(ii) = floor( lower_band(ii)/tick_size ) * tick_size ;
mid_band(ii) = ( upper_band(ii) + lower_band(ii) ) / 2.0 ;
} // end of main loop
retval_list(2) = mid_band ;
retval_list(1) = lower_band ;
retval_list(0) = upper_band ;
return retval_list ;
} // end of function
```

Due to formatting issues on Blogger none of the code boxes contain the required header statements, which should be:-#include octave/oct.h

#include octave/dColVector.h

#include math.h

with the customary < before "octave" and "math" and > after ".h"

Subscribe to:
Posts (Atom)