Facing issue AttributeError: 'DataFrame' object has no attribute 'to_delayed' for dask-xgboost

[bhavana3]

What happened: Facing an issue like AttributeError: 'DataFrame' object has no attribute 'to_delayed' from core.py (line 175, version latest - 0.1.11) data_parts = data.to_delayed() What you expected to happen: train model by calling bst = dxgb.train(client, hyperparameters, df, labels)

Minimal Complete Verifiable Example:

# 
import dask.dataframe as dd

df = dd.read_csv(data_files_location + files_pattern, compression='gzip', delimiter=CONTROL_A_DELIM, quoting=csv.QUOTE_NONE, header=0,
                         na_values=NA_DELIM, keep_default_na=False, dtype=str, blocksize=None)
labels = df[label]
df.drop(label, axis=1, inplace=True)
bst = dxgb.train(client, hyperparameters, df, labels)

Environment:

• Dask version: >=1.15.2
• Python version:3.7
• Operating System: N/A (using jupyter notebooks on cloud)
• Install method (conda, pip, source): conda