- 1.28.0 (latest)
- 1.27.0
- 1.26.0
- 1.25.0
- 1.24.0
- 1.22.0
- 1.21.0
- 1.20.0
- 1.19.0
- 1.18.0
- 1.17.0
- 1.16.0
- 1.15.0
- 1.14.0
- 1.13.0
- 1.12.0
- 1.11.1
- 1.10.0
- 1.9.0
- 1.8.0
- 1.7.0
- 1.6.0
- 1.5.0
- 1.4.0
- 1.3.0
- 1.2.0
- 1.1.0
- 1.0.0
- 0.26.0
- 0.25.0
- 0.24.0
- 0.23.0
- 0.22.0
- 0.21.0
- 0.20.1
- 0.19.2
- 0.18.0
- 0.17.0
- 0.16.0
- 0.15.0
- 0.14.1
- 0.13.0
- 0.12.0
- 0.11.0
- 0.10.0
- 0.9.0
- 0.8.0
- 0.7.0
- 0.6.0
- 0.5.0
- 0.4.0
- 0.3.0
- 0.2.0
API documentation for metrics
package.
Modules
pairwise
API documentation for pairwise
module.
Packages Functions
accuracy_score
accuracy_score(
y_true: typing.Union[bigframes.dataframe.DataFrame, bigframes.series.Series],
y_pred: typing.Union[bigframes.dataframe.DataFrame, bigframes.series.Series],
*,
normalize=True
) -> float
Accuracy classification score.
Examples:
>>> import bigframes.pandas as bpd
>>> import bigframes.ml.metrics
>>> bpd.options.display.progress_bar = None
>>> y_true = bpd.DataFrame([0, 2, 1, 3])
>>> y_pred = bpd.DataFrame([0, 1, 2, 3])
>>> accuracy_score = bigframes.ml.metrics.accuracy_score(y_true, y_pred)
>>> accuracy_score
0.5
If False, return the number of correctly classified samples:
>>> accuracy_score = bigframes.ml.metrics.accuracy_score(y_true, y_pred, normalize=False)
>>> accuracy_score
2
Parameters | |
---|---|
Name | Description |
y_true |
Ground truth (correct) labels. |
y_pred |
Predicted labels, as returned by a classifier. |
normalize |
Default to True. If |
auc
auc(
x: typing.Union[bigframes.dataframe.DataFrame, bigframes.series.Series],
y: typing.Union[bigframes.dataframe.DataFrame, bigframes.series.Series],
) -> float
Compute Area Under the Curve (AUC) using the trapezoidal rule.
This is a general function, given points on a curve. For computing the
area under the ROC-curve, see roc_auc_score
. For an alternative
way to summarize a precision-recall curve, see
average_precision_score
.
Examples:
>>> import bigframes.pandas as bpd
>>> import bigframes.ml.metrics
>>> bpd.options.display.progress_bar = None
>>> x = bpd.DataFrame([1, 1, 2, 2])
>>> y = bpd.DataFrame([2, 3, 4, 5])
>>> auc = bigframes.ml.metrics.auc(x, y)
>>> auc
3.5
The input can be Series:
>>> df = bpd.DataFrame(
... {"x": [1, 1, 2, 2],
... "y": [2, 3, 4, 5],}
... )
>>> auc = bigframes.ml.metrics.auc(df["x"], df["y"])
>>> auc
3.5
Parameters | |
---|---|
Name | Description |
x |
X coordinates. These must be either monotonic increasing or monotonic decreasing. |
y |
Y coordinates. |
confusion_matrix
confusion_matrix(
y_true: typing.Union[bigframes.dataframe.DataFrame, bigframes.series.Series],
y_pred: typing.Union[bigframes.dataframe.DataFrame, bigframes.series.Series],
) -> pandas.core.frame.DataFrame
Compute confusion matrix to evaluate the accuracy of a classification.
By definition a confusion matrix :math:C
is such that :math:C_{i, j}
is equal to the number of observations known to be in group :math:i
and
predicted to be in group :math:j
.
Thus in binary classification, the count of true negatives is
:math:C_{0,0}
, false negatives is :math:C_{1,0}
, true positives is
:math:C_{1,1}
and false positives is :math:C_{0,1}
.
Examples:
>>> import bigframes.pandas as bpd
>>> import bigframes.ml.metrics
>>> bpd.options.display.progress_bar = None
>>> y_true = bpd.DataFrame([2, 0, 2, 2, 0, 1])
>>> y_pred = bpd.DataFrame([0, 0, 2, 2, 0, 2])
>>> confusion_matrix = bigframes.ml.metrics.confusion_matrix(y_true, y_pred)
>>> confusion_matrix
0 1 2
0 2 0 0
1 0 0 1
2 1 0 2
>>> y_true = bpd.DataFrame(["cat", "ant", "cat", "cat", "ant", "bird"])
>>> y_pred = bpd.DataFrame(["ant", "ant", "cat", "cat", "ant", "cat"])
>>> confusion_matrix = bigframes.ml.metrics.confusion_matrix(y_true, y_pred)
>>> confusion_matrix
ant bird cat
ant 2 0 0
bird 0 0 1
cat 1 0 2
Parameters | |
---|---|
Name | Description |
y_true |
Ground truth (correct) target values. |
y_pred |
Estimated targets as returned by a classifier. |
f1_score
f1_score(
y_true: typing.Union[bigframes.dataframe.DataFrame, bigframes.series.Series],
y_pred: typing.Union[bigframes.dataframe.DataFrame, bigframes.series.Series],
*,
average: str = "binary"
) -> pandas.core.series.Series
Compute the F1 score, also known as balanced F-score or F-measure.
The F1 score can be interpreted as a harmonic mean of the precision and recall, where an F1 score reaches its best value at 1 and worst score at 0. The relative contribution of precision and recall to the F1 score are equal. The formula for the F1 score is: F1 = 2 * (precision * recall) / (precision + recall).
In the multi-class and multi-label case, this is the average of
the F1 score of each class with weighting depending on the average
parameter.
Examples:
>>> import bigframes.pandas as bpd
>>> import bigframes.ml.metrics
>>> bpd.options.display.progress_bar = None
>>> y_true = bpd.DataFrame([0, 1, 2, 0, 1, 2])
>>> y_pred = bpd.DataFrame([0, 2, 1, 0, 0, 1])
>>> f1_score = bigframes.ml.metrics.f1_score(y_true, y_pred, average=None)
>>> f1_score
0 0.8
1 0.0
2 0.0
dtype: float64
Parameters | |
---|---|
Name | Description |
y_true |
Series or DataFrame of shape (n_samples,). Ground truth (correct) target values. |
y_pred |
Series or DataFrame of shape (n_samples,). Estimated targets as returned by a classifier. |
mean_squared_error
mean_squared_error(
y_true: typing.Union[bigframes.dataframe.DataFrame, bigframes.series.Series],
y_pred: typing.Union[bigframes.dataframe.DataFrame, bigframes.series.Series],
) -> float
Mean squared error regression loss.
Examples:
>>> import bigframes.pandas as bpd
>>> import bigframes.ml.metrics
>>> bpd.options.display.progress_bar = None
>>> y_true = bpd.DataFrame([3, -0.5, 2, 7])
>>> y_pred = bpd.DataFrame([2.5, 0.0, 2, 8])
>>> mse = bigframes.ml.metrics.mean_squared_error(y_true, y_pred)
>>> mse
0.375
Parameters | |
---|---|
Name | Description |
y_true |
Ground truth (correct) target values. |
y_pred |
Estimated target values. |
precision_score
precision_score(
y_true: typing.Union[bigframes.dataframe.DataFrame, bigframes.series.Series],
y_pred: typing.Union[bigframes.dataframe.DataFrame, bigframes.series.Series],
*,
average: str = "binary"
) -> pandas.core.series.Series
Compute the precision.
The precision is the ratio tp / (tp + fp)
, where tp
is the number of
true positives and fp
the number of false positives. The precision is
intuitively the ability of the classifier not to label as positive a sample
that is negative.
The best value is 1 and the worst value is 0.
Examples:
>>> import bigframes.pandas as bpd
>>> import bigframes.ml.metrics
>>> bpd.options.display.progress_bar = None
>>> y_true = bpd.DataFrame([0, 1, 2, 0, 1, 2])
>>> y_pred = bpd.DataFrame([0, 2, 1, 0, 0, 1])
>>> precision_score = bigframes.ml.metrics.precision_score(y_true, y_pred, average=None)
>>> precision_score
0 0.666667
1 0.000000
2 0.000000
dtype: float64
Parameters | |
---|---|
Name | Description |
y_true |
Series or DataFrame of shape (n_samples,) Ground truth (correct) target values. |
y_pred |
Series or DataFrame of shape (n_samples,) Estimated targets as returned by a classifier. |
r2_score
r2_score(
y_true: typing.Union[bigframes.dataframe.DataFrame, bigframes.series.Series],
y_pred: typing.Union[bigframes.dataframe.DataFrame, bigframes.series.Series],
*,
force_finite=True
) -> float
:math:R^2
(coefficient of determination) regression score function.
Best possible score is 1.0 and it can be negative (because the
model can be arbitrarily worse). In the general case when the true y is
non-constant, a constant model that always predicts the average y
disregarding the input features would get a :math:R^2
score of 0.0.
In the particular case when y_true
is constant, the :math:R^2
score
is not finite: it is either NaN
(perfect predictions) or -Inf
(imperfect predictions). To prevent such non-finite numbers to pollute
higher-level experiments such as a grid search cross-validation, by default
these cases are replaced with 1.0 (perfect predictions) or 0.0 (imperfect
predictions) respectively.
Examples:
>>> import bigframes.pandas as bpd
>>> import bigframes.ml.metrics
>>> bpd.options.display.progress_bar = None
>>> y_true = bpd.DataFrame([3, -0.5, 2, 7])
>>> y_pred = bpd.DataFrame([2.5, 0.0, 2, 8])
>>> r2_score = bigframes.ml.metrics.r2_score(y_true, y_pred)
>>> r2_score
0.9486081370449679
Parameters | |
---|---|
Name | Description |
y_true |
Ground truth (correct) target values. |
y_pred |
Estimated target values. |
recall_score
recall_score(
y_true: typing.Union[bigframes.dataframe.DataFrame, bigframes.series.Series],
y_pred: typing.Union[bigframes.dataframe.DataFrame, bigframes.series.Series],
*,
average: str = "binary"
) -> pandas.core.series.Series
Compute the recall.
The recall is the ratio tp / (tp + fn)
, where tp
is the number of
true positives and fn
the number of false negatives. The recall is
intuitively the ability of the classifier to find all the positive samples.
The best value is 1 and the worst value is 0.
Examples:
>>> import bigframes.pandas as bpd
>>> import bigframes.ml.metrics
>>> bpd.options.display.progress_bar = None
>>> y_true = bpd.DataFrame([0, 1, 2, 0, 1, 2])
>>> y_pred = bpd.DataFrame([0, 2, 1, 0, 0, 1])
>>> recall_score = bigframes.ml.metrics.recall_score(y_true, y_pred, average=None)
>>> recall_score
0 1
1 0
2 0
dtype: int64
Parameters | |
---|---|
Name | Description |
y_true |
Ground truth (correct) target values. |
y_pred |
Estimated targets as returned by a classifier. |
average |
This parameter is required for multiclass/multilabel targets. Possible values are 'None', 'micro', 'macro', 'samples', 'weighted', 'binary'. Only average=None is supported. |
roc_auc_score
roc_auc_score(
y_true: typing.Union[bigframes.dataframe.DataFrame, bigframes.series.Series],
y_score: typing.Union[bigframes.dataframe.DataFrame, bigframes.series.Series],
) -> float
Compute Area Under the Receiver Operating Characteristic Curve (ROC AUC) from prediction scores.
Examples:
>>> import bigframes.pandas as bpd
>>> import bigframes.ml.metrics
>>> bpd.options.display.progress_bar = None
>>> y_true = bpd.DataFrame([0, 0, 1, 1, 0, 1, 0, 1, 1, 1])
>>> y_score = bpd.DataFrame([0.1, 0.4, 0.35, 0.8, 0.65, 0.9, 0.5, 0.3, 0.6, 0.45])
>>> roc_auc_score = bigframes.ml.metrics.roc_auc_score(y_true, y_score)
>>> roc_auc_score
0.625
The input can be Series:
>>> df = bpd.DataFrame(
... {"y_true": [0, 0, 1, 1, 0, 1, 0, 1, 1, 1],
... "y_score": [0.1, 0.4, 0.35, 0.8, 0.65, 0.9, 0.5, 0.3, 0.6, 0.45],}
... )
>>> roc_auc_score = bigframes.ml.metrics.roc_auc_score(df["y_true"], df["y_score"])
>>> roc_auc_score
0.625
Parameters | |
---|---|
Name | Description |
y_true |
True labels or binary label indicators. The binary and multiclass cases expect labels with shape (n_samples,) while the multilabel case expects binary label indicators with shape (n_samples, n_classes). |
y_score |
Target scores. * In the binary case, it corresponds to an array of shape |
roc_curve
roc_curve(
y_true: typing.Union[bigframes.dataframe.DataFrame, bigframes.series.Series],
y_score: typing.Union[bigframes.dataframe.DataFrame, bigframes.series.Series],
*,
drop_intermediate: bool = True
) -> typing.Tuple[
bigframes.series.Series, bigframes.series.Series, bigframes.series.Series
]
Compute Receiver operating characteristic (ROC).
Examples:
>>> import bigframes.pandas as bpd
>>> import bigframes.ml.metrics
>>> bpd.options.display.progress_bar = None
>>> y_true = bpd.DataFrame([1, 1, 2, 2])
>>> y_score = bpd.DataFrame([0.1, 0.4, 0.35, 0.8])
>>> fpr, tpr, thresholds = bigframes.ml.metrics.roc_curve(y_true, y_score, drop_intermediate=False)
>>> fpr
0 0.0
1 0.0
2 0.0
3 0.0
4 0.0
Name: fpr, dtype: Float64
>>> tpr
0 0.0
1 0.333333
2 0.5
3 0.833333
4 1.0
Name: tpr, dtype: Float64
>>> thresholds
0 inf
1 0.8
2 0.4
3 0.35
4 0.1
Name: thresholds, dtype: Float64
Parameters | |
---|---|
Name | Description |
y_true |
Series or DataFrame of shape (n_samples,) True binary labels. If labels are not either {-1, 1} or {0, 1}, then pos_label should be explicitly given. |
y_score |
Series or DataFrame of shape (n_samples,) Target scores, can either be probability estimates of the positive class, confidence values, or non-thresholded measure of decisions (as returned by "decision_function" on some classifiers). |