tall_book.q creates entries with incorrect price
Closed this issue · 3 comments
psaris commented
given the first few rows of a cme file:
$ zcat xcme_md_es_fut_20180903-r-00019.gz | grep ESU8 |head
1128=9�9=175�35=X�49=CME�75=20180903�34=1743�52=20180902214503167161015�60=20180902214503166879423�5799=10000100�268=1�279=1�269=0�48=57287�55=ESU8�83=173�270=290575.0�271=25�346=6�1023=2�10=147�
1128=9�9=177�35=X�49=CME�75=20180903�34=1755�52=20180902214503853254318�60=20180902214503853004477�5799=10000100�268=1�279=1�269=0�48=57287�55=ESU8�83=174�270=290375.0�271=46�346=11�1023=10�10=248�
1128=9�9=176�35=X�49=CME�75=20180903�34=1795�52=20180902214506492119601�60=20180902214506491718105�5799=10000100�268=1�279=1�269=0�48=57287�55=ESU8�83=175�270=290400.0�271=57�346=12�1023=9�10=202�
wide_book.q creates this correctly:
date sym time bprice bsize aprice asize msgseq rptseq matchevent underlying
-------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------------
2018.09.03 ESU8 2018.09.02D21:45:03.166000000 ,2905.75 ,25f `float$() `float$() 1743 173 84 ES
2018.09.03 ESU8 2018.09.02D21:45:03.853000000 2905.75 2903.75 25 46f `float$() `float$() 1755 174 84 ES
2018.09.03 ESU8 2018.09.02D21:45:06.491000000 2905.75 2904 2903.75 25 57 46f `float$() `float$() 1795 175 84 ES
specifically, three different price levels appear
tall_book.q, however, adds three rows into the table all with the same level, orders, size and price:
date sym time side level orders size price msgseq rptseq matchevent
-----------------------------------------------------------------------------------------------------
2018.09.03 ESU8 2018.09.02D21:45:03.166000000 BID 2 6 25 290575 1743 173 84
2018.09.03 ESU8 2018.09.02D21:45:03.853000000 BID 2 6 25 290575 1755 174 84
2018.09.03 ESU8 2018.09.02D21:45:06.491000000 BID 2 6 25 290575 1795 175 84
frenzy21 commented
Will resolve.
I think it was an issue in widebook at one stage and the fix never got carried over potentially.