Comments (4)
what about a project like this one?
https://github.com/GFariasR/jaybeams/projects
(second one)
from jaybeams.
This bug report is just pointing out that the output from itch5bookdepth does not match its description. I may be missing something, I think the project you link does not address this problem. It is proposing something new, which might be fun (the main criteria in the jaybeams context), and/or even useful. Can you explain how it does solve the problem in more detail? Are you suggesting that simply removing the itch5bookdepth program and writing the new program would answer the same questions? Or something else?
from jaybeams.
this is what I intended to document. Is this what you saw?
The project is to create a new program, based on:
- ITCH 5 Statistics: itch5stats.cpp
- ITCH 5 Inside Statistics: itch5inside.cpp
- ITCH 5 Book Depth Statistics: itch5bookdepth.cpp
that generates statistics about price range and book depth.
Statistics are built to test the following hypothesis:
A book of prices has two blocks separated by a large empty price range. We
call real prices to the block near the inside and tail prices to the block
further from the inside.
We intend to validate this assumption, as well as define a good heuristic
to assign value to a threshold price to split these 2 blocks.
Under this assumption a book should look like:
{Px_REAL_1 .. Px_REAL_N } ----- {Px_TAIL_1 .. Px_TAIL_M }
- Px_REAL_N < Px_THRESHOLD < Px_TAIL_1
Initially, Px_THRESHOLD will be set as a movil value:
Px_THRESHOLD = Px_REAL_1 + ( k_ticks * 0.01 )
: if the symbol is quoted above (or at) $1.00
or
Px_THRESHOLD = Px_REAL_1 + ( k_ticks * 0.0001 )
: if the symbol is quoted below $1.00
We will try k_ticks = 10000 to start with. That gives us a range of $100
(or $1). It is configuration parameter thou.
There is a discontinuity when the price crosses that $1.00. Nonetheless we
will try to keep the things moving. The following set of statistics are
generated:
- Bid_Real_Price_Inside: Px_REAL_1
- Bid_Real_Price_Diff: Px_REAL_N - Px_REAL_1
- Bid_Real_Price_Depth: N
- Bid_Tail_Price_Min: Px_TAIL_1
- Bid_Tail_Price_Diff: Px_TAIL_M - Px_TAIL_1
- Bid_Tail_Price_Depth: M
- and the same six values of the offer side.
- Statistics are generated per symbol and aggregated.
On Thu, Nov 3, 2016 at 5:55 PM, coryan [email protected] wrote:
This bug report is just pointing out that the output from itch5bookdepth
does not match its description. I may be missing something, I think the
project you link does not address this problem. It is proposing something
new, which might be fun (the main criteria in the jaybeams context), and/or
even useful. Can you explain how it does solve the problem in more detail?
Are you suggesting that simply removing the itch5bookdepth program and
writing the new program would answer the same questions? Or something else?—
You are receiving this because you commented.
Reply to this email directly, view it on GitHub
#30 (comment), or mute
the thread
https://github.com/notifications/unsubscribe-auth/AVd_gbDS6gJASECo9GuRTD-nEmpQq4tNks5q6lhggaJpZM4KoT-u
.
Gabriel Farias
+1 (843) 737 3827
from jaybeams.
For what is worth, the new data shows that there are millions of levels in the book, the overall results across all symbols are:
Name | NSamples | minBookDepth | p25BookDepth | p50BookDepth | p75BookDepth | p90BookDepth | p99BookDepth | p999BookDepth | p9999BookDepth | maxBookDepth |
---|---|---|---|---|---|---|---|---|---|---|
aggregate | 489064030 | 0 | 3912 | 3469864 | 11739980 | 16702050 | 19679292 | 19977016 | 20006788 | 20010097 |
the book is really deep indeed (most of that is from the SELL side I suspect). In contrast, look at the depth of the events:
Name | NSamples | minBookDepth | p25BookDepth | p50BookDepth | p75BookDepth | p90BookDepth | p99BookDepth | p999BookDepth | p9999BookDepth | maxBookDepth |
---|---|---|---|---|---|---|---|---|---|---|
aggregate | 489064030 | 0 | 0 | 1 | 6 | 14 | 203 | 2135 | 15489331 | 20009799 |
This confirms our approach to just keeping the top N levels in an array and the rest in a more sparse data structure.
ITCH.20150202.bookdepth.csv.txt
ITCH.20150202.eventdepth.csv.txt
from jaybeams.
Related Issues (20)
- Write down guidelines for build matrix HOT 2
- Setup at least one Travis build using the Gallium libraries
- Move images to the jaybeams organization in hub.docker.com
- Only create runtime images from builds with -O2 or -O3.
- Fix ugly hack to disable 'AMD SUMO' GPU.
- Capture image id in benchmark executions
- Create development image before compiling the code HOT 1
- TDE multi array benchmark requires further investigation HOT 2
- Remove hack in ut_conjugate_and_multiply when pocl-0.13 is no longer supported.
- jb::cpu_set is Linux specific
- Support dynamically sized cpu sets in jb::cpu_set
- Protobuf build rules require GNU Make HOT 1
- Complicated rules to build protobuf files in directory hierarchies.
- Rename unit tests from ut_* to *_ut
- Do not ignore coverage count for unit tests
- Move submodules to ext/ subdirectory. HOT 1
- Create class to observe leader election HOT 1
- Create issues for all TODO() entries in jb/etcd HOT 1
- Use leader election in itch-5.x feed handler.
- Migrate all mocks from Skye to GoogleMock HOT 1
Recommend Projects
-
React
A declarative, efficient, and flexible JavaScript library for building user interfaces.
-
Vue.js
🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
-
Typescript
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
-
TensorFlow
An Open Source Machine Learning Framework for Everyone
-
Django
The Web framework for perfectionists with deadlines.
-
Laravel
A PHP framework for web artisans
-
D3
Bring data to life with SVG, Canvas and HTML. 📊📈🎉
-
Recommend Topics
-
javascript
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
-
web
Some thing interesting about web. New door for the world.
-
server
A server is a program made to process requests and deliver data to clients.
-
Machine learning
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
-
Visualization
Some thing interesting about visualization, use data art
-
Game
Some thing interesting about game, make everyone happy.
Recommend Org
-
Facebook
We are working to build community through open source technology. NB: members must have two-factor auth.
-
Microsoft
Open source projects and samples from Microsoft.
-
Google
Google ❤️ Open Source for everyone.
-
Alibaba
Alibaba Open Source for everyone
-
D3
Data-Driven Documents codes.
-
Tencent
China tencent open source team.
from jaybeams.