Skip to content

Commit 3722316

Browse files
jtlaitnmaarnio
jtlait
authored and
nmaarnio
committed
Revert "Remove check for simplex sample space from coda transformations."
This reverts commit abea0f8.
1 parent 49c6917 commit 3722316

File tree

7 files changed

+605
-600
lines changed

7 files changed

+605
-600
lines changed

‎eis_toolkit/transformations/coda/alr.py

+88-88
Original file line numberDiff line numberDiff line change
@@ -1,88 +1,88 @@
1-
from numbers import Number
2-
3-
import numpy as np
4-
import pandas as pd
5-
from beartype import beartype
6-
from beartype.typing import Optional, Sequence
7-
8-
from eis_toolkit.exceptions import InvalidColumnException, NumericValueSignException
9-
from eis_toolkit.utilities.aitchison_geometry import _closure
10-
from eis_toolkit.utilities.checks.compositional import check_compositional_data
11-
from eis_toolkit.utilities.miscellaneous import rename_columns_by_pattern
12-
13-
14-
@beartype
15-
def _alr_transform(df: pd.DataFrame, columns: Sequence[str], denominator_column: str) -> pd.DataFrame:
16-
17-
ratios = df[columns].div(df[denominator_column], axis=0)
18-
return np.log(ratios)
19-
20-
21-
@beartype
22-
def alr_transform(
23-
df: pd.DataFrame, column: Optional[str] = None, keep_denominator_column: bool = False
24-
) -> pd.DataFrame:
25-
"""
26-
Perform an additive logratio transformation on the data.
27-
28-
Args:
29-
df: A dataframe of compositional data.
30-
column: The name of the column to be used as the denominator column.
31-
keep_denominator_column: Whether to include the denominator column in the result. If True, the returned
32-
dataframe retains its original shape.
33-
34-
Returns:
35-
A new dataframe containing the ALR transformed data.
36-
37-
Raises:
38-
InvalidColumnException: The input column isn't found in the dataframe.
39-
InvalidCompositionException: Data is not normalized to the expected value.
40-
NumericValueSignException: Data contains zeros or negative values.
41-
"""
42-
check_compositional_data(df)
43-
44-
if column is not None and column not in df.columns:
45-
raise InvalidColumnException(f"The column {column} was not found in the dataframe.")
46-
47-
column = column if column is not None else df.columns[-1]
48-
49-
columns = [col for col in df.columns]
50-
51-
if not keep_denominator_column and column in columns:
52-
columns.remove(column)
53-
54-
return rename_columns_by_pattern(_alr_transform(df, columns, column))
55-
56-
57-
@beartype
58-
def _inverse_alr(df: pd.DataFrame, denominator_column: str, scale: Number = 1.0) -> pd.DataFrame:
59-
dfc = df.copy()
60-
61-
if denominator_column not in dfc.columns.values:
62-
# Add the denominator column
63-
dfc[denominator_column] = 0.0
64-
65-
return _closure(np.exp(dfc), scale)
66-
67-
68-
@beartype
69-
def inverse_alr(df: pd.DataFrame, denominator_column: str, scale: Number = 1.0) -> pd.DataFrame:
70-
"""
71-
Perform the inverse transformation for a set of ALR transformed data.
72-
73-
Args:
74-
df: A dataframe of ALR transformed compositional data.
75-
denominator_column: The name of the denominator column.
76-
scale: The value to which each composition should be normalized. Eg., if the composition is expressed
77-
as percentages, scale=100.
78-
79-
Returns:
80-
A dataframe containing the inverse transformed data.
81-
82-
Raises:
83-
NumericValueSignException: The input scale value is zero or less.
84-
"""
85-
if scale <= 0:
86-
raise NumericValueSignException("The scale value should be positive.")
87-
88-
return _inverse_alr(df, denominator_column, scale)
1+
from numbers import Number
2+
3+
import numpy as np
4+
import pandas as pd
5+
from beartype import beartype
6+
from beartype.typing import Optional, Sequence
7+
8+
from eis_toolkit.exceptions import InvalidColumnException, NumericValueSignException
9+
from eis_toolkit.utilities.aitchison_geometry import _closure
10+
from eis_toolkit.utilities.checks.compositional import check_in_simplex_sample_space
11+
from eis_toolkit.utilities.miscellaneous import rename_columns_by_pattern
12+
13+
14+
@beartype
15+
def _alr_transform(df: pd.DataFrame, columns: Sequence[str], denominator_column: str) -> pd.DataFrame:
16+
17+
ratios = df[columns].div(df[denominator_column], axis=0)
18+
return np.log(ratios)
19+
20+
21+
@beartype
22+
def alr_transform(
23+
df: pd.DataFrame, column: Optional[str] = None, keep_denominator_column: bool = False
24+
) -> pd.DataFrame:
25+
"""
26+
Perform an additive logratio transformation on the data.
27+
28+
Args:
29+
df: A dataframe of compositional data.
30+
column: The name of the column to be used as the denominator column.
31+
keep_denominator_column: Whether to include the denominator column in the result. If True, the returned
32+
dataframe retains its original shape.
33+
34+
Returns:
35+
A new dataframe containing the ALR transformed data.
36+
37+
Raises:
38+
InvalidColumnException: The input column isn't found in the dataframe.
39+
InvalidCompositionException: Data is not normalized to the expected value.
40+
NumericValueSignException: Data contains zeros or negative values.
41+
"""
42+
check_in_simplex_sample_space(df)
43+
44+
if column is not None and column not in df.columns:
45+
raise InvalidColumnException(f"The column {column} was not found in the dataframe.")
46+
47+
column = column if column is not None else df.columns[-1]
48+
49+
columns = [col for col in df.columns]
50+
51+
if not keep_denominator_column and column in columns:
52+
columns.remove(column)
53+
54+
return rename_columns_by_pattern(_alr_transform(df, columns, column))
55+
56+
57+
@beartype
58+
def _inverse_alr(df: pd.DataFrame, denominator_column: str, scale: Number = 1.0) -> pd.DataFrame:
59+
dfc = df.copy()
60+
61+
if denominator_column not in dfc.columns.values:
62+
# Add the denominator column
63+
dfc[denominator_column] = 0.0
64+
65+
return _closure(np.exp(dfc), scale)
66+
67+
68+
@beartype
69+
def inverse_alr(df: pd.DataFrame, denominator_column: str, scale: Number = 1.0) -> pd.DataFrame:
70+
"""
71+
Perform the inverse transformation for a set of ALR transformed data.
72+
73+
Args:
74+
df: A dataframe of ALR transformed compositional data.
75+
denominator_column: The name of the denominator column.
76+
scale: The value to which each composition should be normalized. Eg., if the composition is expressed
77+
as percentages, scale=100.
78+
79+
Returns:
80+
A dataframe containing the inverse transformed data.
81+
82+
Raises:
83+
NumericValueSignException: The input scale value is zero or less.
84+
"""
85+
if scale <= 0:
86+
raise NumericValueSignException("The scale value should be positive.")
87+
88+
return _inverse_alr(df, denominator_column, scale)

‎eis_toolkit/transformations/coda/clr.py

+79-79
Original file line numberDiff line numberDiff line change
@@ -1,79 +1,79 @@
1-
from numbers import Number
2-
3-
import numpy as np
4-
import pandas as pd
5-
from beartype import beartype
6-
from beartype.typing import Optional, Sequence
7-
from scipy.stats import gmean
8-
9-
from eis_toolkit.exceptions import NumericValueSignException
10-
from eis_toolkit.utilities.aitchison_geometry import _closure
11-
from eis_toolkit.utilities.checks.compositional import check_compositional_data
12-
from eis_toolkit.utilities.miscellaneous import rename_columns, rename_columns_by_pattern
13-
14-
15-
@beartype
16-
def _centered_ratio(row: pd.Series) -> pd.Series:
17-
18-
return row / gmean(row)
19-
20-
21-
@beartype
22-
def _clr_transform(df: pd.DataFrame) -> pd.DataFrame:
23-
24-
dfc = df.copy()
25-
dfc = dfc.apply(_centered_ratio, axis=1)
26-
27-
return np.log(dfc)
28-
29-
30-
@beartype
31-
def clr_transform(df: pd.DataFrame) -> pd.DataFrame:
32-
"""
33-
Perform a centered logratio transformation on the data.
34-
35-
Args:
36-
df: A dataframe of compositional data.
37-
38-
Returns:
39-
A new dataframe containing the CLR transformed data.
40-
41-
Raises:
42-
InvalidCompositionException: Data is not normalized to the expected value.
43-
NumericValueSignException: Data contains zeros or negative values.
44-
"""
45-
check_compositional_data(df)
46-
return rename_columns_by_pattern(_clr_transform(df))
47-
48-
49-
@beartype
50-
def _inverse_clr(df: pd.DataFrame, colnames: Optional[Sequence[str]] = None, scale: Number = 1.0) -> pd.DataFrame:
51-
inverse = _closure(np.exp(df), scale)
52-
53-
if colnames is not None:
54-
return rename_columns(inverse, colnames)
55-
56-
return inverse
57-
58-
59-
@beartype
60-
def inverse_clr(df: pd.DataFrame, colnames: Optional[Sequence[str]] = None, scale: Number = 1.0) -> pd.DataFrame:
61-
"""
62-
Perform the inverse transformation for a set of CLR transformed data.
63-
64-
Args:
65-
df: A dataframe of CLR transformed compositional data.
66-
colnames: List of column names to rename the columns to.
67-
scale: The value to which each composition should be normalized. Eg., if the composition is expressed
68-
as percentages, scale=100.
69-
70-
Returns:
71-
A dataframe containing the inverse transformed data.
72-
73-
Raises:
74-
NumericValueSignException: The input scale value is zero or less.
75-
"""
76-
if scale <= 0:
77-
raise NumericValueSignException("The scale value should be positive.")
78-
79-
return _inverse_clr(df, colnames, scale)
1+
from numbers import Number
2+
3+
import numpy as np
4+
import pandas as pd
5+
from beartype import beartype
6+
from beartype.typing import Optional, Sequence
7+
from scipy.stats import gmean
8+
9+
from eis_toolkit.exceptions import NumericValueSignException
10+
from eis_toolkit.utilities.aitchison_geometry import _closure
11+
from eis_toolkit.utilities.checks.compositional import check_in_simplex_sample_space
12+
from eis_toolkit.utilities.miscellaneous import rename_columns, rename_columns_by_pattern
13+
14+
15+
@beartype
16+
def _centered_ratio(row: pd.Series) -> pd.Series:
17+
18+
return row / gmean(row)
19+
20+
21+
@beartype
22+
def _clr_transform(df: pd.DataFrame) -> pd.DataFrame:
23+
24+
dfc = df.copy()
25+
dfc = dfc.apply(_centered_ratio, axis=1)
26+
27+
return np.log(dfc)
28+
29+
30+
@beartype
31+
def clr_transform(df: pd.DataFrame) -> pd.DataFrame:
32+
"""
33+
Perform a centered logratio transformation on the data.
34+
35+
Args:
36+
df: A dataframe of compositional data.
37+
38+
Returns:
39+
A new dataframe containing the CLR transformed data.
40+
41+
Raises:
42+
InvalidCompositionException: Data is not normalized to the expected value.
43+
NumericValueSignException: Data contains zeros or negative values.
44+
"""
45+
check_in_simplex_sample_space(df)
46+
return rename_columns_by_pattern(_clr_transform(df))
47+
48+
49+
@beartype
50+
def _inverse_clr(df: pd.DataFrame, colnames: Optional[Sequence[str]] = None, scale: Number = 1.0) -> pd.DataFrame:
51+
inverse = _closure(np.exp(df), scale)
52+
53+
if colnames is not None:
54+
return rename_columns(inverse, colnames)
55+
56+
return inverse
57+
58+
59+
@beartype
60+
def inverse_clr(df: pd.DataFrame, colnames: Optional[Sequence[str]] = None, scale: Number = 1.0) -> pd.DataFrame:
61+
"""
62+
Perform the inverse transformation for a set of CLR transformed data.
63+
64+
Args:
65+
df: A dataframe of CLR transformed compositional data.
66+
colnames: List of column names to rename the columns to.
67+
scale: The value to which each composition should be normalized. Eg., if the composition is expressed
68+
as percentages, scale=100.
69+
70+
Returns:
71+
A dataframe containing the inverse transformed data.
72+
73+
Raises:
74+
NumericValueSignException: The input scale value is zero or less.
75+
"""
76+
if scale <= 0:
77+
raise NumericValueSignException("The scale value should be positive.")
78+
79+
return _inverse_clr(df, colnames, scale)

0 commit comments

Comments
 (0)