# Copyright 2022 The SQLNet Company GmbH
#
# This file is licensed under the Elastic License 2.0 (ELv2).
# Refer to the LICENSE.txt file in the root of the repository
# for details.
#
"""
Container for data frames in memory.
"""
from getml.data import DataFrame
from getml.data.helpers import list_data_frames
from getml.data.helpers2 import load_data_frame
from getml.utilities.formatting import _Formatter
# --------------------------------------------------------------------
[docs]class DataFrames:
"""
Container which holds all data frames associated with the running
project that are currently stored in memory. The container supports
slicing and is sort- and filterable.
"""
# ----------------------------------------------------------------
def __init__(self, data=None):
self._in_memory = list_data_frames()["in_memory"]
self._on_disk = list_data_frames()["on_disk"]
if data is None:
self.data = [load_data_frame(name) for name in self._in_memory]
else:
self.data = data
# ----------------------------------------------------------------
def __getitem__(self, key):
if isinstance(key, int):
return self.data[key]
if isinstance(key, slice):
dfs_subset = self.data[key]
return DataFrames(data=dfs_subset)
if isinstance(key, str):
if key in self.in_memory:
return [df for df in self.data if df.name == key][0]
if key in self.on_disk:
raise AttributeError(f"DataFrame {key} not loaded from disk.")
raise AttributeError(f"No DataFrame with name: {key}")
raise TypeError(
f"DataFrames can only be indexed by: int, slices, or str, not {type(key).__name__}"
)
# ----------------------------------------------------------------
def __len__(self):
return len(self.data)
# ----------------------------------------------------------------
def __repr__(self):
if len(self.in_memory) == 0:
output = "No data frames in memory."
else:
output = self._format()._render_string()
if len(self.on_disk) > 0:
output += "\n\nOn disk:\n"
output += "\n".join(self.on_disk)
return output
# ----------------------------------------------------------------
def _repr_html_(self):
if len(self.in_memory) == 0:
output = "<p>No data frames in memory.</p>"
else:
output = self._format()._render_html()
if len(self.on_disk) > 0:
output += "<p>On disk:</p>"
output += "<br>".join(self.on_disk)
return output
# ----------------------------------------------------------------
def _format(self):
headers = [["name", "rows", "columns", "memory usage"]]
rows = [[df.name, df.nrows(), df.ncols(), df.memory_usage] for df in self.data]
formatted = _Formatter(headers, rows)
formatted[4].cell_template = "{:{width}.2f} MB"
return formatted
# ----------------------------------------------------------------
[docs] def delete(self):
"""
Deletes all data frames in the current project.
Args:
mem_only (bool):
If called with the `mem_only` option set to True, the data
frames will be kept on disk (in the project folder) and can
be reloaded to memory through
:meth:`getml.project.data_frames.load_all`.
"""
for name in self.on_disk:
DataFrame(name).delete()
# ----------------------------------------------------------------
@property
def in_memory(self):
"""
Returns the names of all data frames currently in memory.
"""
return self._in_memory
# ----------------------------------------------------------------
[docs] def filter(self, conditional):
"""
Filters the data frames container.
Args:
conditional (callable):
A callable that evaluates to a boolean for a given item.
Returns:
:class:`getml.pipeline.DataFrames`:
A container of filtered data frames.
Example:
.. code-block:: python
big_frames = getml.project.data_frames.filter(lambda frame: frame.memory_usage > 1000)
"""
dfs_filtered = [df for df in self.data if conditional(df)]
return DataFrames(data=dfs_filtered)
# ----------------------------------------------------------------
[docs] def load(self):
"""
Loads all data frames stored in the project folder to memory.
"""
for df in self.on_disk:
if df not in self.in_memory:
self.data.append(load_data_frame(df))
# ----------------------------------------------------------------
@property
def on_disk(self):
"""
Returns the names of all data frames stored in the project folder.
"""
return self._on_disk
# ----------------------------------------------------------------
[docs] def retrieve(self):
"""
Retrieve a dict of all data frames in memory.
"""
return {df.name: df for df in self.data}
# ----------------------------------------------------------------
[docs] def save(self):
"""
Saves all data frames currently in memory to disk.
"""
for df in self.data:
df.save()
# ----------------------------------------------------------------
[docs] def sort(self, key, descending=False):
"""
Sorts the data frames container.
Args:
key (callable, optional):
A callable that evaluates to a sort key for a given item.
descending (bool, optional):
Whether to sort in descending order.
Return:
:class:`getml.pipeline.DataFrames`:
A container of sorted data frames.
Example:
.. code-block:: python
by_num_rows = getml.project.data_frames.sort(lambda frame: frame.nrows())
"""
dfs_sorted = sorted(self.data, key=key, reverse=descending)
return DataFrames(data=dfs_sorted)
# ----------------------------------------------------------------
[docs] def unload(self):
"""
Unloads all data frames in the current project from memory.
"""
for name in self.on_disk:
DataFrame(name).unload()