def __getitem__(self, key):
is_valid, msg, target_axis = is_valid_index(key, self.axis)
if not is_valid:
print_warning(
"{} Pnguin auto-handles this, but this could take a second for larger datasets".format(
msg
),
self.warnings,
)
return format_input(self.data, target_axis)[key]
return fetch_item(self.data, key)
def head(self, n: int = 5):
"""Return the head (or the top n rows) of your DataFrame
Args:
n (int, optional): The number of rows to be returned. 5 by default
Returns:
(pnguin.DataFrame): A new DataFrame with the first n rows of self
"""
if self.axis == Axis.row:
return DataFrame(self.data[0:n], self.axis)
row_data = format_input(self.data, Axis.row)
return DataFrame(row_data[0:n], self.axis)
def __init__(self, data, axis: Axis = Axis.col, warnings: bool = True):
"""Create a new pnguin DataFrame
Args:
data (any): Input data for the pnguin dataframe, either in list of dicts (rows) or dict of lists (columns) form
axis (str, optional): The primary axis that the DataFrame is to operate using. 'col' by default
warnings (bool, optional): A flag determining whether pnguin will print warnings. True by default
Returns:
(pnguin.DataFrame): A new pnguin DataFrame
"""
self.axis = axis
self.data = format_input(data, axis)
self.warnings = warnings
def apply(self, x: Callable, axis: Axis = Axis.row, inplace: bool = False):
"""Apply a given function per row or per column on a DataFrame
Args:
x (callable): A function to be run per row/column
axis (str): The target axis to be operated on ('row' by default)
inplace (bool): Determines whether the existing DataFrame will be modified, or a new DataFrame will be returned
Returns:
(pnguin.DataFrame or None)
"""
target = self._data_as_rows() if axis == Axis.row else self._data_as_cols()
applied = apply_rows(target, x) if axis == Axis.row else apply_cols(target, x)
if inplace:
self.data = format_input(applied, self.axis)
return None
return DataFrame(applied, self.axis)
def __setitem__(self, key, value):
is_valid, msg, target_axis = is_valid_index(key, self.axis)
if not is_valid:
print_warning(
"{} Pnguin auto-handles this, but this could take a second for larger datasets".format(
msg
),
self.warnings,
)
data = (
self._data_as_cols()
if target_axis == Axis.col
else self._data_as_rows()
)
set_item(data, key, value)
self.data = format_input(data, self.axis)
else:
set_item(self.data, key, value)
def dropna(self, exclude: list = [], inplace: bool = False):
"""Drop those rows with NaN values
Args:
exclude (list, optional): Certain column names that may be excempt from the dropping criteria. [] by default
inplace (bool, optional): Perform this operation in-place, and don't return a new DataFrame. False by default
Returns:
(pnguin.DataFrame or None)
"""
target = self._data_as_rows()
rectified = drop_nan_rows(target, exclude)
if inplace:
self.data = format_input(rectified, self.axis)
return None
return DataFrame(rectified, self.axis)
def _data_as_cols(self):
return self.data if self.axis == Axis.col else format_input(self.data, Axis.col)
def _data_as_rows(self):
return self.data if self.axis == Axis.row else format_input(self.data, Axis.row)