Skip to content

tea_tasting.utils #

Useful functions and classes.

check_scalar(value, name='value', *, typ=None, ge=None, gt=None, le=None, lt=None, ne=None, in_=None) #

Check if a scalar parameter meets specified type and value constraints.

Parameters:

Name Type Description Default
value R

Parameter value.

 required
name str

Parameter name.

 'value'
typ object

Acceptable data types.

 None
ge object

If not None, check that parameter value is greater than or equal to ge.

 None
gt object

If not None, check that parameter value is greater than gt.

 None
le object

If not None, check that parameter value is less than or equal to le.

 None
lt object

If not None, check that parameter value is less than lt.

 None
ne object

If not None, check that parameter value is not equal to ne.

 None
in_ object

If not None, check that parameter value is in in_.

 None

Returns:

Type Description
R

Parameter value.
Source code in src/tea_tasting/utils.py 
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
def check_scalar(  # noqa: PLR0913
    value: R,
    name: str = "value",
    *,
    typ: object = None,
    ge: object = None,
    gt: object = None,
    le: object = None,
    lt: object = None,
    ne: object = None,
    in_: object = None,
) -> R:
    """Check if a scalar parameter meets specified type and value constraints.

    Args:
        value: Parameter value.
        name: Parameter name.
        typ: Acceptable data types.
        ge: If not `None`, check that parameter value is greater than
            or equal to `ge`.
        gt: If not `None`, check that parameter value is greater than `gt`.
        le: If not `None`, check that parameter value is less than or equal to `le`.
        lt: If not `None`, check that parameter value is less than `lt`.
        ne: If not `None`, check that parameter value is not equal to `ne`.
        in_: If not `None`, check that parameter value is in `in_`.

    Returns:
        Parameter value.
    """
    if typ is not None and not isinstance(value, typ):  # type: ignore
        raise TypeError(f"{name} must be an instance of {typ}.")
    if ge is not None and value < ge:
        raise ValueError(f"{name} == {value}, must be >= {ge}.")
    if gt is not None and value <= gt:
        raise ValueError(f"{name} == {value}, must be > {gt}.")
    if le is not None and value > le:
        raise ValueError(f"{name} == {value}, must be <= {le}.")
    if lt is not None and value >= lt:
        raise ValueError(f"{name} == {value}, must be < {lt}.")
    if ne is not None and value == ne:
        raise ValueError(f"{name} == {value}, must be != {ne}.")
    if in_ is not None and value not in in_:
        raise ValueError(f"{name} == {value}, must be in {in_}.")
    return value

auto_check(value, name) #

Automatically check a parameter's type and value based on its name.

The following parameter names are supported: "alpha", "alternative", "confidence_level", "correction", "equal_var", "n_obs", "n_resamples", "power", "ratio", "use_t".

Parameters:

Name Type Description Default
value R

Parameter value.

 required
name str

Parameter name.

 required

Returns:

Type Description
R

Parameter value.
Source code in src/tea_tasting/utils.py 
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
def auto_check(value: R, name: str) -> R:  # noqa: C901, PLR0912
    """Automatically check a parameter's type and value based on its name.

    The following parameter names are supported: `"alpha"`, `"alternative"`,
    `"confidence_level"`, `"correction"`, `"equal_var"`, `"n_obs"`,
    `"n_resamples"`, `"power"`, `"ratio"`, `"use_t"`.

    Args:
        value: Parameter value.
        name: Parameter name.

    Returns:
        Parameter value.
    """
    if name == "alpha":
        check_scalar(value, name, typ=float, gt=0, lt=1)
    if name == "alternative":
        check_scalar(value, name, typ=str, in_={"two-sided", "greater", "less"})
    elif name == "confidence_level":
        check_scalar(value, name, typ=float, gt=0, lt=1)
    elif name == "correction":
        check_scalar(value, name, typ=bool)
    elif name == "equal_var":
        check_scalar(value, name, typ=bool)
    elif name == "n_obs":
        check_scalar(value, name, typ=int | Sequence | None)
        if isinstance(value, int):
            check_scalar(value, name, gt=1)
        if isinstance(value, Sequence):
            for val in value:
                check_scalar(val, name, typ=int, gt=1)
    elif name == "n_resamples":
        check_scalar(value, name, typ=int, gt=0)
    elif name == "power":
        check_scalar(value, name, typ=float, gt=0, lt=1)
    elif name == "ratio":
        check_scalar(value, name, typ=float | int, gt=0)
    elif name == "use_t":
        check_scalar(value, name, typ=bool)
    return value

format_num(val, sig=3, *, pct=False, nan='-', inf='∞', fixed_point_range=(0.001, 10000000), thousands_sep=None, decimal_point=None) #

Format a number according to specified formatting rules.

Parameters:

Name Type Description Default
val float | int | None

Number to format.

 required
sig int

Number of significant digits.

 3
pct bool

If True, format as a percentage.

 False
nan str

Replacement for None and nan values.

 '-'
inf str

Replacement for infinite values.

 '∞'
fixed_point_range tuple[float | None, float | None]

The range within which the number is formatted as fixed point. Number outside of the range is formatted as exponential. None means no boundary.

 (0.001, 10000000)
thousands_sep str | None

Thousands separator. If None, the value from locales is used.

 None
decimal_point str | None

Decimal point symbol. If None, the value from locales is used.

 None

Returns:

Type Description
str

Formatted number.
Source code in src/tea_tasting/utils.py 
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
def format_num(
    val: float | int | None,
    sig: int = 3,
    *,
    pct: bool = False,
    nan: str = "-",
    inf: str = "∞",
    fixed_point_range: tuple[float | None, float | None] = (0.001, 10_000_000),
    thousands_sep: str | None = None,
    decimal_point: str | None = None,
) -> str:
    """Format a number according to specified formatting rules.

    Args:
        val: Number to format.
        sig: Number of significant digits.
        pct: If `True`, format as a percentage.
        nan: Replacement for `None` and `nan` values.
        inf: Replacement for infinite values.
        fixed_point_range: The range within which the number is formatted
            as fixed point.
            Number outside of the range is formatted as exponential.
            `None` means no boundary.
        thousands_sep: Thousands separator. If `None`, the value from locales is used.
        decimal_point: Decimal point symbol. If `None`, the value from locales is used.

    Returns:
        Formatted number.
    """
    if val is None or math.isnan(val):
        return nan

    if math.isinf(val):
        return inf if val > 0 else "-" + inf

    if pct:
        val = val * 100

    if (
        (fixed_point_range[0] is not None and abs(val) < fixed_point_range[0]) or
        (fixed_point_range[1] is not None and abs(val) >= fixed_point_range[1])
    ):
        precision = max(0, sig - 1)
        typ = "e" if val != 0 else "f"
    else:
        precision = max(0, sig - 1 - int(math.floor(math.log10(abs(val)))))
        val = round(val, precision)
        # Repeat in order to format 99.999 as "100", not "100.0".
        precision = max(0, sig - 1 - int(math.floor(math.log10(abs(val)))))
        typ = "f"

    result = format(val, f"_.{precision}{typ}")

    if thousands_sep is None:
        thousands_sep = locale.localeconv().get("thousands_sep", "_")  # type: ignore
    if thousands_sep != "_":
        result = result.replace("_", thousands_sep)

    if decimal_point is None:
        decimal_point = locale.localeconv().get("decimal_point", ".")  # type: ignore
    if decimal_point != ".":
        result = result.replace(".", decimal_point)

    if pct:
        return result + "%"

    return result

get_and_format_num(data, key) #

Get and format dictionary value.

Formatting rules:

  • If a name starts with "rel_" or equals to "power" consider it a percentage value. Round percentage values to 2 significant digits, multiply by 100 and add "%".
  • Round other values to 3 significant values.
  • If value is less than 0.001 or is greater than or equal to 10_000_000, format it in exponential presentation.
  • If a name ends with "_ci", consider it a confidence interval. Look up for attributes "{name}_lower" and "{name}_upper", and format the interval as "[{lower_bound}, {upper_bound}]".

Parameters:

Name Type Description Default
data dict[str, object]

Dictionary.

 required
key str

Key.

 required

Returns:

Type Description
str

Formatted value.
Source code in src/tea_tasting/utils.py 
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
def get_and_format_num(data: dict[str, object], key: str) -> str:
    """Get and format dictionary value.

    Formatting rules:

    - If a name starts with `"rel_"` or equals to `"power"` consider it
        a percentage value. Round percentage values to 2 significant digits,
        multiply by `100` and add `"%"`.
    - Round other values to 3 significant values.
    - If value is less than `0.001` or is greater than or equal to `10_000_000`,
        format it in exponential presentation.
    - If a name ends with `"_ci"`, consider it a confidence interval.
        Look up for attributes `"{name}_lower"` and `"{name}_upper"`,
        and format the interval as `"[{lower_bound}, {upper_bound}]"`.

    Args:
        data: Dictionary.
        key: Key.

    Returns:
        Formatted value.
    """
    if key.endswith("_ci"):
        ci_lower = get_and_format_num(data, key + "_lower")
        ci_upper = get_and_format_num(data, key + "_upper")
        return f"[{ci_lower}, {ci_upper}]"

    val = data.get(key)
    if not isinstance(val, float | int | None):
        return str(val)

    sig, pct = (2, True) if key.startswith("rel_") or key == "power" else (3, False)
    return format_num(val, sig=sig, pct=pct)

DictsReprMixin #

Bases: ABC

Representation and conversion of a sequence of dictionaries.

Default formatting rules:

  • If a name starts with "rel_" or equals to "power" consider it a percentage value. Round percentage values to 2 significant digits, multiply by 100 and add "%".
  • Round other values to 3 significant values.
  • If value is less than 0.001 or is greater than or equal to 10_000_000, format it in exponential presentation.
  • If a name ends with "_ci", consider it a confidence interval. Look up for attributes "{name}_lower" and "{name}_upper", and format the interval as "[{lower_bound}, {upper_bound}]".

to_dicts() abstractmethod #

Convert the object to a sequence of dictionaries.

Source code in src/tea_tasting/utils.py 
294
295
296
@abc.abstractmethod
def to_dicts(self) -> Sequence[dict[str, object]]:
    """Convert the object to a sequence of dictionaries."""

to_arrow() #

Convert the object to a PyArrow Table.

Source code in src/tea_tasting/utils.py 
298
299
300
def to_arrow(self) -> pa.Table:
    """Convert the object to a PyArrow Table."""
    return pa.Table.from_pylist(self.to_dicts())

to_pandas() #

Convert the object to a Pandas DataFrame.

Source code in src/tea_tasting/utils.py 
302
303
304
305
def to_pandas(self) -> PandasDataFrame:
    """Convert the object to a Pandas DataFrame."""
    import pandas as pd
    return pd.DataFrame.from_records(self.to_dicts())

to_polars() #

Convert the object to a Polars DataFrame.

Source code in src/tea_tasting/utils.py 
307
308
309
310
def to_polars(self) -> PolarsDataFrame:
    """Convert the object to a Polars DataFrame."""
    import polars as pl
    return pl.from_dicts(self.to_dicts())

to_pretty_dicts(keys=None, formatter=get_and_format_num) #

Convert the object to a list of dictionaries with formatted values.

Default formatting rules:

  • If a name starts with "rel_" or equals to "power" consider it a percentage value. Round percentage values to 2 significant digits, multiply by 100 and add "%".
  • Round other values to 3 significant values.
  • If value is less than 0.001 or is greater than or equal to 10_000_000, format it in exponential presentation.
  • If a name ends with "_ci", consider it a confidence interval. Look up for attributes "{name}_lower" and "{name}_upper", and format the interval as "[{lower_bound}, {upper_bound}]".

Parameters:

Name Type Description Default
keys Sequence[str] | None

Keys to convert. If a key is not defined in the dictionary it's assumed to be None.

 None
formatter Callable[[dict[str, object], str], str]

Custom formatter function. It should accept a dictionary of metric result attributes and an attribute name, and return a formatted attribute value.

 get_and_format_num

Returns:

Type Description
list[dict[str, str]]

List of dictionaries with formatted values.
Source code in src/tea_tasting/utils.py 
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
def to_pretty_dicts(
    self,
    keys: Sequence[str] | None = None,
    formatter: Callable[[dict[str, object], str], str] = get_and_format_num,
) -> list[dict[str, str]]:
    """Convert the object to a list of dictionaries with formatted values.

    Default formatting rules:

    - If a name starts with `"rel_"` or equals to `"power"` consider it
        a percentage value. Round percentage values to 2 significant digits,
        multiply by `100` and add `"%"`.
    - Round other values to 3 significant values.
    - If value is less than `0.001` or is greater than or equal to `10_000_000`,
        format it in exponential presentation.
    - If a name ends with `"_ci"`, consider it a confidence interval.
        Look up for attributes `"{name}_lower"` and `"{name}_upper"`,
        and format the interval as `"[{lower_bound}, {upper_bound}]"`.

    Args:
        keys: Keys to convert. If a key is not defined in the dictionary
            it's assumed to be `None`.
        formatter: Custom formatter function. It should accept a dictionary
            of metric result attributes and an attribute name, and return
            a formatted attribute value.

    Returns:
        List of dictionaries with formatted values.
    """
    if keys is None:
        keys = self.default_keys
    return [{key: formatter(data, key) for key in keys} for data in self.to_dicts()]

to_string(keys=None, formatter=get_and_format_num) #

Convert the object to a string.

Default formatting rules:

  • If a name starts with "rel_" or equals to "power" consider it a percentage value. Round percentage values to 2 significant digits, multiply by 100 and add "%".
  • Round other values to 3 significant values.
  • If value is less than 0.001 or is greater than or equal to 10_000_000, format it in exponential presentation.
  • If a name ends with "_ci", consider it a confidence interval. Look up for attributes "{name}_lower" and "{name}_upper", and format the interval as "[{lower_bound}, {upper_bound}]".

Parameters:

Name Type Description Default
keys Sequence[str] | None

Keys to convert. If a key is not defined in the dictionary it's assumed to be None.

 None
formatter Callable[[dict[str, object], str], str]

Custom formatter function. It should accept a dictionary of metric result attributes and an attribute name, and return a formatted attribute value.

 get_and_format_num

Returns:

Type Description
str

A table with results rendered as string.
Source code in src/tea_tasting/utils.py 
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
def to_string(
    self,
    keys: Sequence[str] | None = None,
    formatter: Callable[[dict[str, object], str], str] = get_and_format_num,
) -> str:
    """Convert the object to a string.

    Default formatting rules:

    - If a name starts with `"rel_"` or equals to `"power"` consider it
        a percentage value. Round percentage values to 2 significant digits,
        multiply by `100` and add `"%"`.
    - Round other values to 3 significant values.
    - If value is less than `0.001` or is greater than or equal to `10_000_000`,
        format it in exponential presentation.
    - If a name ends with `"_ci"`, consider it a confidence interval.
        Look up for attributes `"{name}_lower"` and `"{name}_upper"`,
        and format the interval as `"[{lower_bound}, {upper_bound}]"`.

    Args:
        keys: Keys to convert. If a key is not defined in the dictionary
            it's assumed to be `None`.
        formatter: Custom formatter function. It should accept a dictionary
            of metric result attributes and an attribute name, and return
            a formatted attribute value.

    Returns:
        A table with results rendered as string.
    """
    if keys is None:
        keys = self.default_keys
    widths = {key: len(key) for key in keys}

    pretty_dicts = []
    for data in self.to_dicts():
        pretty_dict = {}
        for key in keys:
            val = formatter(data, key)
            widths[key] = max(widths[key], len(val))
            pretty_dict |= {key: val}
        pretty_dicts.append(pretty_dict)

    sep = " "
    rows = [sep.join(key.rjust(widths[key]) for key in keys)]
    rows.extend(
        sep.join(pretty_dict[key].rjust(widths[key]) for key in keys)
        for pretty_dict in pretty_dicts
    )
    return "\n".join(rows)

to_html(keys=None, formatter=get_and_format_num, *, indent=None) #

Convert the object to HTML.

Default formatting rules:

  • If a name starts with "rel_" or equals to "power" consider it a percentage value. Round percentage values to 2 significant digits, multiply by 100 and add "%".
  • Round other values to 3 significant values.
  • If value is less than 0.001 or is greater than or equal to 10_000_000, format it in exponential presentation.
  • If a name ends with "_ci", consider it a confidence interval. Look up for attributes "{name}_lower" and "{name}_upper", and format the interval as "[{lower_bound}, {upper_bound}]".

Parameters:

Name Type Description Default
keys Sequence[str] | None

Keys to convert. If a key is not defined in the dictionary it's assumed to be None.

 None
formatter Callable[[dict[str, object], str], str]

Custom formatter function. It should accept a dictionary of metric result attributes and an attribute name, and return a formatted attribute value.

 get_and_format_num
indent str | None

Whitespace to insert for each indentation level. If None, do not indent.

 None

Returns:

Type Description
str

A table with results rendered as HTML.
Source code in src/tea_tasting/utils.py 
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
def to_html(
    self,
    keys: Sequence[str] | None = None,
    formatter: Callable[[dict[str, object], str], str] = get_and_format_num,
    *,
    indent: str | None = None,
) -> str:
    """Convert the object to HTML.

    Default formatting rules:

    - If a name starts with `"rel_"` or equals to `"power"` consider it
        a percentage value. Round percentage values to 2 significant digits,
        multiply by `100` and add `"%"`.
    - Round other values to 3 significant values.
    - If value is less than `0.001` or is greater than or equal to `10_000_000`,
        format it in exponential presentation.
    - If a name ends with `"_ci"`, consider it a confidence interval.
        Look up for attributes `"{name}_lower"` and `"{name}_upper"`,
        and format the interval as `"[{lower_bound}, {upper_bound}]"`.

    Args:
        keys: Keys to convert. If a key is not defined in the dictionary
            it's assumed to be `None`.
        formatter: Custom formatter function. It should accept a dictionary
            of metric result attributes and an attribute name, and return
            a formatted attribute value.
        indent: Whitespace to insert for each indentation level. If `None`,
            do not indent.

    Returns:
        A table with results rendered as HTML.
    """
    if keys is None:
        keys = self.default_keys
    table = ET.Element(
        "table",
        {"class": "dataframe", "style": "text-align: right;"},
    )
    thead = ET.SubElement(table, "thead")
    thead_tr = ET.SubElement(thead, "tr")
    for key in keys:
        th = ET.SubElement(thead_tr, "th")
        th.text = key
    tbody = ET.SubElement(table, "tbody")
    for data in self.to_dicts():
        tr = ET.SubElement(tbody, "tr")
        for key in keys:
            td = ET.SubElement(tr, "td")
            td.text = formatter(data, key)
    if indent is not None:
        ET.indent(table, space=indent)
    return ET.tostring(table, encoding="unicode", method="html")

ReprMixin #

A mixin class that provides a method for generating a string representation.

Representation string is generated based on parameters values saved in attributes.

div(numer, denom, fill_zero_div='auto') #

Perform division, providing specified results for cases of division by zero.

Parameters:

Name Type Description Default
numer float | int

Numerator.

 required
denom float | int

Denominator.

 required
fill_zero_div float | int | Literal['auto']

Result if denominator is zero.

 'auto'

Returns:

Type Description
float | int

Result of the division.

If fill_zero_div is equal "auto", return:

  • inf if numerator is greater than 0,
  • nan if numerator is equal to or less than 0.
Source code in src/tea_tasting/utils.py 
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
def div(
    numer: float | int,
    denom: float | int,
    fill_zero_div: float | int | Literal["auto"] = "auto",
) -> float |int:
    """Perform division, providing specified results for cases of division by zero.

    Args:
        numer: Numerator.
        denom: Denominator.
        fill_zero_div: Result if denominator is zero.

    Returns:
        Result of the division.

    If `fill_zero_div` is equal `"auto"`, return:

    - `inf` if numerator is greater than `0`,
    - `nan` if numerator is equal to or less than `0`.
    """
    if denom != 0:
        return numer / denom
    if fill_zero_div != "auto":
        return fill_zero_div
    return float("inf") if numer > 0 else float("nan")

Float #

Bases: _NumericBase, float

Float that gracefully handles division by zero errors.

Int #

Bases: _NumericBase, int

Integer that gracefully handles division by zero errors.

numeric(value, fill_zero_div='auto') #

Float or integer that gracefully handles division by zero errors.

Source code in src/tea_tasting/utils.py 
639
640
641
642
643
644
645
646
647
648
649
650
651
def numeric(
    value: object,
    fill_zero_div: float | int | Literal["auto"] = "auto",
) -> Numeric:
    """Float or integer that gracefully handles division by zero errors."""
    if isinstance(value, int):
        return Int(value, fill_zero_div)
    if isinstance(value, float):
        return Float(value, fill_zero_div)
    try:
        return Int(value, fill_zero_div)
    except ValueError:
        return Float(value, fill_zero_div)