oceanprotocol / pdr-backend

Instructions & code to run predictoors, traders, more.
Apache License 2.0
22 stars 15 forks source link

[Aimodel, PPSS] aimodel_data_factory ERROR Too little data -> Need better ppss defaults #1260

Closed trentmc closed 1 week ago

trentmc commented 1 week ago

The bug / to reproduce

With defaults, in console:

pdr sim ppss.yaml

I get:

2024-06-20 12:16:23,860 aimodel_data_factory ERROR Too little data. To fix:broaden time, or shrink testshift, max_diff, or autoregr_n

Full traceback

(venv) trentmc@trents-m3: ~/code/pdr-backend $ pdr sim ppss.yaml 
do_lake_submcommand: parsed_args: ['sim', 'ppss.yaml']
2024-06-20 12:16:16,055 cli INFO pdr sim: Begin
2024-06-20 12:16:16,055 cli INFO Arguments:
2024-06-20 12:16:16,055 cli INFO PPSS_FILE=ppss.yaml
2024-06-20 12:16:16,055 cli INFO Nested args: {}
2024-06-20 12:16:16,074 cli WARNING Multiple predict feeds provided, using the first one
2024-06-20 12:16:16,075 sim_engine INFO Start run
2024-06-20 12:16:16,076 sim_engine INFO Initialize plot data.
2024-06-20 12:16:16,076 ohlcv_data_factory INFO Get historical data, across many exchanges & pairs: begin.
2024-06-20 12:16:16,077 ohlcv_data_factory INFO Data start: timestamp=1717323376077, dt=2024-06-02_10:16:16.077
2024-06-20 12:16:16,077 ohlcv_data_factory INFO Data fin: timestamp=1718878576076, dt=2024-06-20_10:16:16.076
2024-06-20 12:16:16,077 ohlcv_data_factory INFO Update all rawohlcv files: begin
2024-06-20 12:16:16,077 ohlcv_data_factory INFO Update rawohlcv file at exch=binance, pair=BTC/USDT: begin
2024-06-20 12:16:16,077 ohlcv_data_factory INFO filename=/Users/trentmc/code/pdr-backend/lake_data/binance_BTC-USDT_5m.parquet
2024-06-20 12:16:16,077 ohlcv_data_factory INFO File already exists
2024-06-20 12:16:16,082 ohlcv_data_factory INFO File starts at: timestamp=1687183800000, dt=2023-06-19_14:10:00.000
2024-06-20 12:16:16,082 ohlcv_data_factory INFO File finishes at: timestamp=1718877000000, dt=2024-06-20_09:50:00.000
2024-06-20 12:16:16,082 ohlcv_data_factory INFO User-specified start >= file start, so append file
2024-06-20 12:16:16,082 ohlcv_data_factory INFO Aim to fetch data from start time: timestamp=1718877300000, dt=2024-06-20_09:55:00.000
2024-06-20 12:16:16,083 ohlcv_data_factory INFO Fetch up to 1000 pts from timestamp=1718877300000, dt=2024-06-20_09:55:00.000
2024-06-20 12:16:19,926 lake_plutil INFO Just appended 5 df rows to file /Users/trentmc/code/pdr-backend/lake_data/binance_BTC-USDT_5m.parquet
2024-06-20 12:16:19,926 ohlcv_data_factory INFO Update rawohlcv file at exch=binance, pair=BTC/USDT: done
2024-06-20 12:16:19,926 ohlcv_data_factory INFO Update rawohlcv file at exch=binance, pair=ETH/USDT: begin
2024-06-20 12:16:19,926 ohlcv_data_factory INFO filename=/Users/trentmc/code/pdr-backend/lake_data/binance_ETH-USDT_5m.parquet
2024-06-20 12:16:19,926 ohlcv_data_factory INFO File already exists
2024-06-20 12:16:19,929 ohlcv_data_factory INFO File starts at: timestamp=1703328900000, dt=2023-12-23_10:55:00.000
2024-06-20 12:16:19,929 ohlcv_data_factory INFO File finishes at: timestamp=1718877000000, dt=2024-06-20_09:50:00.000
2024-06-20 12:16:19,930 ohlcv_data_factory INFO User-specified start >= file start, so append file
2024-06-20 12:16:19,930 ohlcv_data_factory INFO Aim to fetch data from start time: timestamp=1718877300000, dt=2024-06-20_09:55:00.000
2024-06-20 12:16:19,930 ohlcv_data_factory INFO Fetch up to 1000 pts from timestamp=1718877300000, dt=2024-06-20_09:55:00.000
2024-06-20 12:16:23,821 lake_plutil INFO Just appended 5 df rows to file /Users/trentmc/code/pdr-backend/lake_data/binance_ETH-USDT_5m.parquet
2024-06-20 12:16:23,821 ohlcv_data_factory INFO Update rawohlcv file at exch=binance, pair=ETH/USDT: done
2024-06-20 12:16:23,821 ohlcv_data_factory INFO Update all rawohlcv files: done
2024-06-20 12:16:23,821 ohlcv_data_factory INFO Load rawohlcv file.
2024-06-20 12:16:23,860 ohlcv_data_factory INFO Get historical data, across many exchanges & pairs: done.
2024-06-20 12:16:23,860 aimodel_data_factory ERROR Too little data. To fix:broaden time, or shrink testshift, max_diff, or autoregr_n