Docs

BDS+BEEP

I already have raw cycler files. How do I use BDS before a BEEP workflow?

BDS+BEEP

Question

I already have raw cycler files. How do I use BDS before a BEEP workflow?

Use This When

Use this path when you want BDS to normalize or screen raw cycler exports before they enter a BEEP structuring or feature-generation pipeline.

BEEP has its own supported raw-file formats and structuring logic. BDS should not pretend to be a native BEEP replacement. The useful bridge is:

Let BDS detect, validate, and normalize messy raw files.
Use the BDS report to decide whether a file is safe for a BEEP pipeline.
For formats BEEP does not natively structure, write a BEEP-style staging CSV and a metadata sidecar for a small local BEEP adapter.

Convert Raw Files

bds convert raw.mpt beep_staging.csv \
  --target beep \
  --current-sign preserve \
  --repair-policy repair \
  --report normalized.report.json

Create A BEEP-Style Staging CSV

import json
import polars as pl
 
beep_staging = pl.read_csv("beep_staging.csv")
 
metadata = {
    "source": "battery-data-standard",
    "bds_report": "normalized.report.json",
    "current_sign": "preserve",
    "units": {
        "test_time": "s",
        "voltage": "V",
        "current": "A",
        "charge_capacity": "Ah",
        "discharge_capacity": "Ah",
    },
}
 
with open("beep_staging.metadata.json", "w", encoding="utf-8") as handle:
    json.dump(metadata, handle, indent=2)

Suggested Mapping

BDS export column	BEEP-style staging column
`Test Time (s)`	`test_time`
`Voltage (V)`	`voltage`
`Current (A)`	`current`
`Cycle Count`	`cycle_index`
`Step Index`	`step_index`
`Step Time (s)`	`step_time`
`Charging Capacity (Ah)`	`charge_capacity`
`Discharging Capacity (Ah)`	`discharge_capacity`

Batch Screening Before BEEP

bds batch raw_exports bds_exports \
  --recursive \
  --target beep \
  --format parquet \
  --manifest bds_manifest.jsonl \
  --current-sign preserve \
  --repair-policy repair

Use bds_manifest.jsonl to separate:

converted time-series files;
EIS files, which should not be passed to a time-series BEEP pipeline;
unsupported helper files;
conversion errors that need a new fixture.

Known Limits

This page does not claim that BEEP can directly ingest every BDS export.
When BEEP already supports a raw cycler file, prefer BEEP's native structuring path and use BDS as an independent validation/preflight step.
For raw formats only BDS supports, keep the BDS staging CSV plus metadata and implement a small local BEEP adapter around that normalized table.