tiled.adapters.dataframe.DataFrameAdapter

class tiled.adapters.dataframe.DataFrameAdapter(partitions, meta, divisions, *, metadata=None, specs=None, references=None)

Wrap a dataframe-like object in an interface that Tiled can serve.

Examples

>>> df = pandas.DataFrame({"a": [1, 2, 3], "b": [4, 5, 6]})
>>> DataFrameAdapter(dask.dataframe.from_pandas(df), npartitions=1)

Read a CSV (uses dask.dataframe.read_csv).

>>> DataFrameAdapter.read_csv("myfiles.*.csv")
>>> DataFrameAdapter.read_csv("s3://bucket/myfiles.*.csv")

__init__(partitions, meta, divisions, *, metadata=None, specs=None, references=None)

Methods

__init__(partitions, meta, divisions, *[, ...])
from_dask_dataframe(ddf[, metadata, specs, ...])
from_pandas(*args[, metadata, specs, references])
items()
macrostructure()
microstructure()
read([fields])
read_csv(*args[, metadata, specs, references])	Read a CSV.
read_partition(partition[, fields])

Attributes

metadata
structure_family