def count_missing_data():
ebola = ds.load_ebola()
print('Eboda data:', '\n', ebola.head(), '\n')
print('# of entries per column:', '\n', ebola.count(), '\n')
num_rows = ebola.shape[0]
num_missing = num_rows - ebola.count()
print('# of missing entries per column:', '\n', num_missing, '\n')
def backward_fill_missing_values():
ebola = ds.load_ebola()
# Fills missing values with the NEXT valid observation. In case the previous row is NaN, it doesn't do anything.
# Observe the column 'Cases_Liberia' and the rows 2 and 3
print('Ebola data:', '\n',
ebola.fillna(method='bfill').ix[0:, 0:5].head(), '\n')
print('Ebola data:', '\n',
ebola.fillna(method='bfill').ix[0:, 0:5].tail(), '\n')
def arithmetic_ops_missing_values():
ebola = ds.load_ebola()
ebola['sum_columns'] = ebola['Cases_Guinea'] + ebola[
'Cases_Liberia'] + ebola['Cases_SierraLeone']
ebola_subset = ebola.ix[:, [
'Cases_Guinea', 'Cases_Liberia', 'Cases_SierraLeone'
]]
print('Summing column with NaN values', '\n', ebola_subset.head(n=10),
'\n')
def count_missing_values_numpy():
ebola = ds.load_ebola()
print('Eboda data:', '\n', ebola.head(), '\n')
missing_values = np.count_nonzero(ebola.isnull())
print('Missing values in DataFrame:', '\n', missing_values, '\n')
missing_values = np.count_nonzero(ebola['Cases_Guinea'].isnull())
print('Missing values in Series:', '\n', missing_values, '\n')
# NaN values will be on the top, so look at the head()
missing_values = ebola['Cases_Guinea'].value_counts(dropna=False).head()
print('Missing values in Series II:', '\n', missing_values, '\n')
def fill_missing_values():
ebola = ds.load_ebola()
# Fills missing values with the value on the previous row. In case the previous row is NaN, it doesn't do anything.
# We can also use fillna(0, inplace=True) to change the current DataFrame. It will avoid creating a copy
print('Ebola data:', '\n', ebola.fillna(0).ix[0:10, 0:5], '\n')
def drop_nan_values():
ebola = ds.load_ebola()
# One can also drop rows with NaN values, but that has to be done with caution.
# If we drop the th NaN values form the ebola DataFrame, we will end up with only one row.
# It returns a new DataFrame, unless we use inplace=True
print('Dropping NaN values:', '\n', ebola.dropna().shape)
def forward_fill_missing_values():
ebola = ds.load_ebola()
# Fills values with the vLAST valid observation. In case the previous row is NaN, it doesn't do anything.
# Observe the column 'Cases_Liberia' and the rows 2 and 3
print('Ebola data:', '\n',
ebola.fillna(method='ffill').ix[0:10, 0:5], '\n')