{
  "all_source_statistics_match": true,
  "canonical_all_totals": {
    "both_modality_usable_records": 212440,
    "c_max_peaks": 263,
    "c_null_records": 0,
    "c_total_peaks": 3287622,
    "h_max_peaks": 34,
    "h_null_records": 0,
    "h_total_peaks": 1898067,
    "j_empty_peaks": 1053267,
    "j_nonempty_peaks": 844800,
    "j_null_peaks": 0,
    "rows": 212440,
    "zero_integration_peaks": 0
  },
  "canonical_duplicate_record_ids": 0,
  "canonical_schema_matches": true,
  "dataset": "nmrtrans",
  "source_format": "LZ4-compressed pickle containing NMRTrans/NMRSpec rows",
  "source_observations": {
    "c_shift_outlier_peaks": 444,
    "c_shift_outlier_peaks_by_split": {
      "test": 39,
      "train": 359,
      "val": 46
    },
    "c_shift_outlier_records": 372,
    "c_shift_outlier_records_by_split": {
      "test": 37,
      "train": 297,
      "val": 38
    },
    "h_shift_outlier_peaks": 1332,
    "h_shift_outlier_peaks_by_split": {
      "test": 136,
      "train": 1086,
      "val": 110
    },
    "h_shift_outlier_records": 1182,
    "h_shift_outlier_records_by_split": {
      "test": 126,
      "train": 950,
      "val": 106
    },
    "missing_j_policy": "Raw missing J values become empty canonical lists for this dataset.",
    "raw_missing_j_peaks": 0,
    "raw_missing_j_peaks_by_split": {
      "test": 0,
      "train": 0,
      "val": 0
    },
    "records_over_60_c_peaks": 33,
    "records_over_60_c_peaks_by_split": {
      "test": 5,
      "train": 25,
      "val": 3
    },
    "records_over_60_h_peaks": 0,
    "records_over_60_h_peaks_by_split": {
      "test": 0,
      "train": 0,
      "val": 0
    },
    "shift_outlier_thresholds_ppm": {
      "c": "outside [-20, 300]",
      "h": "outside [-5, 20]"
    },
    "zero_integration_records": 0,
    "zero_integration_records_by_split": {
      "test": 0,
      "train": 0,
      "val": 0
    }
  },
  "source_split_bytes_total": 44247243,
  "source_splits": {
    "test": {
      "bytes": 4430229,
      "path": "models/NMRTrans/data/test.pkl.lz4",
      "rows": 21298
    },
    "train": {
      "bytes": 35400465,
      "path": "models/NMRTrans/data/train.pkl.lz4",
      "rows": 169863
    },
    "val": {
      "bytes": 4416549,
      "path": "models/NMRTrans/data/val.pkl.lz4",
      "rows": 21279
    }
  },
  "source_to_canonical_matches": {
    "both_modality_usable_records": true,
    "c_max_peaks": true,
    "c_null_records": true,
    "c_total_peaks": true,
    "h_max_peaks": true,
    "h_null_records": true,
    "h_total_peaks": true,
    "j_empty_peaks": true,
    "j_nonempty_peaks": true,
    "j_null_peaks": true,
    "rows": true,
    "zero_integration_peaks": true
  },
  "source_totals": {
    "both_modality_usable_records": 212440,
    "c_max_peaks": 263,
    "c_null_records": 0,
    "c_total_peaks": 3287622,
    "h_max_peaks": 34,
    "h_null_records": 0,
    "h_total_peaks": 1898067,
    "j_empty_peaks": 1053267,
    "j_nonempty_peaks": 844800,
    "j_null_peaks": 0,
    "rows": 212440,
    "zero_integration_peaks": 0
  },
  "split_consistency": {
    "all_count": 212440,
    "all_equals_split_union": true,
    "extra_in_all_count": 0,
    "missing_from_all_count": 0,
    "split_counts": {
      "test": 21298,
      "train": 169863,
      "val": 21279
    },
    "split_overlap_count": 0
  }
}