There is a difference between dataminer and normal ranking systems for treatment of ties

DataMiner and API
1

When pulling ranked AvgRec data through “rank/ranks” from the API, an AvgRec of 1 (ie. all buys) gets a ranking of 57 (ie. its treating all ties as the lowest rank).

When pulling ranked AvgRec data from the ranking system on the website, an AvgRec of 1 gets 99.87 (ie. its treating all ties as the highest rank).

2

Hi philjoe. I am having trouble reproducing this issue. Could you please include your DataMiner script? Or send it to me if you dont want to post it here. Just the parts with the settings and the section with AvcRec.

I created a ranking system with only a single factor which was AvgRec. It is set to lower is better, vs Universe.
Using Rank_ranks in the API with that ranking system returns 99.9 for the those with avgrec=1. I tried a Ranks DataMiner script that uses this same ranking system and it also returns 99.9. I created a DataMiner script that defines the ranking system inline and it also returned 99.9

Main:
    Operation: Ranks
    On Error:  Stop

Default Settings:
    PIT Method: Prelim 
    Ranking Method: NAsNegative
    Universe: Prussell 1000
    Start Date: 2021-09-11
    End Date: 2021-09-11 #Optional.
    Columns: composite # ( [ranks] | composite | factor )
    Include Names: false
    Ranking System: 
        Nodes:
            -
                Type: StockFormula
                Rank: Lower
                Formula: AvgRec
3

Change it to:

Ranking Method: NAsNeutral
Columns: factor

4

Hi Philip,
Sorry, but I have not been able to reproduce the AvgRec issue. I made the 2 changes you listed to my DataMiner script from my earlier post and stocks with AvgRec=1 are getting a factor rank of 99.9 as they should in every test I ran.
Could you please send me your DataMiner script that is having the issue? You can email it to me rather than posting it here. If it is something you dont want to share, you can delete other factors (except AvgRec), change the weights, etc.

Also, are you running the latest version of DataMiner from April 2021? If not, you can get it here https://www.dropbox.com/home/p123/Data%20Miner/latest_release

5

I am using the API not the dataminer.

client.rank_ranks({“pitMethod”: “Prelim”,‘rankingSystem’: RankingSystemID, ‘asOfDt’: RunDate, ‘universe’: Universe, “nodeDetails”: ‘factor’})

6

I tried multiple ranking systems and universes with the API script settings you sent in your last post and I get ranks of 99.x for the AvgRec factor in every test. Most likely, there is something in you ranking system or universe that is different from what I have tried. Could you please send me the id of your ranking system and universe? The id is the number in the URL when you are on the ranking system or universe page.

thanks,