| | |
- __builtin__.object
-
- Beta
- Cdf
- Dirichlet
- FixedWidthVariables
- HypothesisTest
- Interpolator
- Pdf
-
- EstimatedPdf
- ExponentialPdf
- NormalPdf
- exceptions.Exception(exceptions.BaseException)
-
- UnimplementedMethodException
- _DictWrapper(__builtin__.object)
-
- Hist
- Pmf
-
- Joint
- Suite
class Beta(__builtin__.object) |
| |
Represents a Beta distribution.
See http://en.wikipedia.org/wiki/Beta_distribution
|
| |
Methods defined here:
- EvalPdf(self, x)
- Evaluates the PDF at x.
- MakeCdf(self, steps=101)
- Returns the CDF of this distribution.
- MakePmf(self, steps=101, label=None)
- Returns a Pmf of this distribution.
Note: Normally, we just evaluate the PDF at a sequence
of points and treat the probability density as a probability
mass.
But if alpha or beta is less than one, we have to be
more careful because the PDF goes to infinity at x=0
and x=1. In that case we evaluate the CDF and compute
differences.
- Mean(self)
- Computes the mean of this distribution.
- Random(self)
- Generates a random variate from this distribution.
- Sample(self, n)
- Generates a random sample from this distribution.
n: int sample size
- Update(self, data)
- Updates a Beta distribution.
data: pair of int (heads, tails)
- __init__(self, alpha=1, beta=1, label=None)
- Initializes a Beta distribution.
Data descriptors defined here:
- __dict__
- dictionary for instance variables (if defined)
- __weakref__
- list of weak references to the object (if defined)
|
class Cdf(__builtin__.object) |
| |
Represents a cumulative distribution function.
Attributes:
xs: sequence of values
ps: sequence of probabilities
label: string used as a graph label.
|
| |
Methods defined here:
- ConfidenceInterval = CredibleInterval(self, percentage=90)
- Copy(self, label=None)
- Returns a copy of this Cdf.
label: string label for the new Cdf
- CredibleInterval(self, percentage=90)
- Computes the central credible interval.
If percentage=90, computes the 90% CI.
Args:
percentage: float between 0 and 100
Returns:
sequence of two floats, low and high
- Items(self)
- Returns a sorted sequence of (value, probability) pairs.
Note: in Python3, returns an iterator.
- MakePmf(self, label=None)
- Makes a Pmf.
- Max(self, k)
- Computes the CDF of the maximum of k selections from this dist.
k: int
returns: new Cdf
- Mean(self)
- Computes the mean of a CDF.
Returns:
float mean
- Percentile(self, p)
- Returns the value that corresponds to percentile p.
Args:
p: number in the range [0, 100]
Returns:
number value
- PercentileRank(self, x)
- Returns the percentile rank of the value x.
x: potential value in the CDF
returns: percentile rank in the range 0 to 100
- Prob(self, x)
- Returns CDF(x), the probability that corresponds to value x.
Args:
x: number
Returns:
float probability
- ProbArray = Probs(self, xs)
- Probs(self, xs)
- Gets probabilities for a sequence of values.
xs: any sequence that can be converted to NumPy array
returns: NumPy array of cumulative probabilities
- Random(self)
- Chooses a random value from this distribution.
- Render(self, **options)
- Generates a sequence of points suitable for plotting.
An empirical CDF is a step function; linear interpolation
can be misleading.
Note: options are ignored
Returns:
tuple of (xs, ps)
- Sample(self, n)
- Generates a random sample from this distribution.
n: int length of the sample
returns: NumPy array
- Scale(self, factor)
- Multiplies the xs by a factor.
factor: what to multiply by
- Shift(self, term)
- Adds a term to the xs.
term: how much to add
- Value(self, p)
- Returns InverseCDF(p), the value that corresponds to probability p.
Args:
p: number in the range [0, 1]
Returns:
number value
- ValueArray(self, ps)
- Returns InverseCDF(p), the value that corresponds to probability p.
Args:
ps: NumPy array of numbers in the range [0, 1]
Returns:
NumPy array of values
- Values(self)
- Returns a sorted list of values.
- __delitem__(self)
- __eq__(self, other)
- __getitem__(self, x)
- __init__(self, obj=None, ps=None, label=None)
- Initializes.
If ps is provided, obj must be the corresponding list of values.
obj: Hist, Pmf, Cdf, Pdf, dict, pandas Series, list of pairs
ps: list of cumulative probabilities
label: string label
- __len__(self)
- __repr__ = __str__(self)
- __setitem__(self)
- __str__(self)
Data descriptors defined here:
- __dict__
- dictionary for instance variables (if defined)
- __weakref__
- list of weak references to the object (if defined)
|
class Dirichlet(__builtin__.object) |
| |
Represents a Dirichlet distribution.
See http://en.wikipedia.org/wiki/Dirichlet_distribution
|
| |
Methods defined here:
- Likelihood(self, data)
- Computes the likelihood of the data.
Selects a random vector of probabilities from this distribution.
Returns: float probability
- LogLikelihood(self, data)
- Computes the log likelihood of the data.
Selects a random vector of probabilities from this distribution.
Returns: float log probability
- MarginalBeta(self, i)
- Computes the marginal distribution of the ith element.
See http://en.wikipedia.org/wiki/Dirichlet_distribution
#Marginal_distributions
i: int
Returns: Beta object
- PredictivePmf(self, xs, label=None)
- Makes a predictive distribution.
xs: values to go into the Pmf
Returns: Pmf that maps from x to the mean prevalence of x
- Random(self)
- Generates a random variate from this distribution.
Returns: normalized vector of fractions
- Update(self, data)
- Updates a Dirichlet distribution.
data: sequence of observations, in order corresponding to params
- __init__(self, n, conc=1, label=None)
- Initializes a Dirichlet distribution.
n: number of dimensions
conc: concentration parameter (smaller yields more concentration)
label: string label
Data descriptors defined here:
- __dict__
- dictionary for instance variables (if defined)
- __weakref__
- list of weak references to the object (if defined)
|
class EstimatedPdf(Pdf) |
| |
Represents a PDF estimated by KDE.
|
| |
- Method resolution order:
- EstimatedPdf
- Pdf
- __builtin__.object
Methods defined here:
- Density(self, xs)
- Evaluates this Pdf at xs.
returns: float or NumPy array of probability density
- GetLinspace(self)
- Get a linspace for plotting.
Returns: numpy array
- __init__(self, sample, label=None)
- Estimates the density function based on a sample.
sample: sequence of data
label: string
- __str__(self)
Methods inherited from Pdf:
- Items(self)
- Generates a sequence of (value, probability) pairs.
- MakePmf(self, **options)
- Makes a discrete version of this Pdf.
options can include
label: string
low: low end of range
high: high end of range
n: number of places to evaluate
Returns: new Pmf
- Render(self, **options)
- Generates a sequence of points suitable for plotting.
If options includes low and high, it must also include n;
in that case the density is evaluated an n locations between
low and high, including both.
If options includes xs, the density is evaluate at those location.
Otherwise, self.GetLinspace is invoked to provide the locations.
Returns:
tuple of (xs, densities)
Data descriptors inherited from Pdf:
- __dict__
- dictionary for instance variables (if defined)
- __weakref__
- list of weak references to the object (if defined)
|
class ExponentialPdf(Pdf) |
| |
Represents the PDF of an exponential distribution.
|
| |
- Method resolution order:
- ExponentialPdf
- Pdf
- __builtin__.object
Methods defined here:
- Density(self, xs)
- Evaluates this Pdf at xs.
xs: scalar or sequence of floats
returns: float or NumPy array of probability density
- GetLinspace(self)
- Get a linspace for plotting.
Returns: numpy array
- __init__(self, lam=1, label=None)
- Constructs an exponential Pdf with given parameter.
lam: rate parameter
label: string
- __str__(self)
Methods inherited from Pdf:
- Items(self)
- Generates a sequence of (value, probability) pairs.
- MakePmf(self, **options)
- Makes a discrete version of this Pdf.
options can include
label: string
low: low end of range
high: high end of range
n: number of places to evaluate
Returns: new Pmf
- Render(self, **options)
- Generates a sequence of points suitable for plotting.
If options includes low and high, it must also include n;
in that case the density is evaluated an n locations between
low and high, including both.
If options includes xs, the density is evaluate at those location.
Otherwise, self.GetLinspace is invoked to provide the locations.
Returns:
tuple of (xs, densities)
Data descriptors inherited from Pdf:
- __dict__
- dictionary for instance variables (if defined)
- __weakref__
- list of weak references to the object (if defined)
|
class FixedWidthVariables(__builtin__.object) |
| |
Represents a set of variables in a fixed width file.
|
| |
Methods defined here:
- ReadFixedWidth(self, filename, **options)
- Reads a fixed width ASCII file.
filename: string filename
returns: DataFrame
- __init__(self, variables, index_base=0)
- Initializes.
variables: DataFrame
index_base: are the indices 0 or 1 based?
Attributes:
colspecs: list of (start, end) index tuples
names: list of string variable names
Data descriptors defined here:
- __dict__
- dictionary for instance variables (if defined)
- __weakref__
- list of weak references to the object (if defined)
|
class Hist(_DictWrapper) |
| |
Represents a histogram, which is a map from values to frequencies.
Values can be any hashable type; frequencies are integer counters.
|
| |
- Method resolution order:
- Hist
- _DictWrapper
- __builtin__.object
Methods defined here:
- Freq(self, x)
- Gets the frequency associated with the value x.
Args:
x: number value
Returns:
int frequency
- Freqs(self, xs)
- Gets frequencies for a sequence of values.
- IsSubset(self, other)
- Checks whether the values in this histogram are a subset of
the values in the given histogram.
- Subtract(self, other)
- Subtracts the values in the given histogram from this histogram.
Methods inherited from _DictWrapper:
- Copy(self, label=None)
- Returns a copy.
Make a shallow copy of d. If you want a deep copy of d,
use copy.deepcopy on the whole object.
label: string label for the new Hist
returns: new _DictWrapper with the same type
- Exp(self, m=None)
- Exponentiates the probabilities.
m: how much to shift the ps before exponentiating
If m is None, normalizes so that the largest prob is 1.
- GetDict(self)
- Gets the dictionary.
- Incr(self, x, term=1)
- Increments the freq/prob associated with the value x.
Args:
x: number value
term: how much to increment by
- Items(self)
- Gets an unsorted sequence of (value, freq/prob) pairs.
- Largest(self, n=10)
- Returns the largest n values, with frequency/probability.
n: number of items to return
- Log(self, m=None)
- Log transforms the probabilities.
Removes values with probability 0.
Normalizes so that the largest logprob is 0.
- MakeCdf(self, label=None)
- Makes a Cdf.
- MaxLike(self)
- Returns the largest frequency/probability in the map.
- Mult(self, x, factor)
- Scales the freq/prob associated with the value x.
Args:
x: number value
factor: how much to multiply by
- Print(self)
- Prints the values and freqs/probs in ascending order.
- Remove(self, x)
- Removes a value.
Throws an exception if the value is not there.
Args:
x: value to remove
- Render(self, **options)
- Generates a sequence of points suitable for plotting.
Note: options are ignored
Returns:
tuple of (sorted value sequence, freq/prob sequence)
- Scale(self, factor)
- Multiplies the values by a factor.
factor: what to multiply by
Returns: new object
- Set(self, x, y=0)
- Sets the freq/prob associated with the value x.
Args:
x: number value
y: number freq or prob
- SetDict(self, d)
- Sets the dictionary.
- Smallest(self, n=10)
- Returns the smallest n values, with frequency/probability.
n: number of items to return
- Total(self)
- Returns the total of the frequencies/probabilities in the map.
- Values(self)
- Gets an unsorted sequence of values.
Note: one source of confusion is that the keys of this
dictionary are the values of the Hist/Pmf, and the
values of the dictionary are frequencies/probabilities.
- __contains__(self, value)
- __delitem__(self, value)
- __eq__(self, other)
- __getitem__(self, value)
- __hash__(self)
- __init__(self, obj=None, label=None)
- Initializes the distribution.
obj: Hist, Pmf, Cdf, Pdf, dict, pandas Series, list of pairs
label: string label
- __iter__(self)
- __len__(self)
- __repr__ = __str__(self)
- __setitem__(self, value, prob)
- __str__(self)
- iterkeys(self)
- Returns an iterator over keys.
Data descriptors inherited from _DictWrapper:
- __dict__
- dictionary for instance variables (if defined)
- __weakref__
- list of weak references to the object (if defined)
|
class HypothesisTest(__builtin__.object) |
| |
Represents a hypothesis test.
|
| |
Methods defined here:
- MakeModel(self)
- Build a model of the null hypothesis.
- MaxTestStat(self)
- Returns the largest test statistic seen during simulations.
- PValue(self, iters=1000)
- Computes the distribution of the test statistic and p-value.
iters: number of iterations
returns: float p-value
- PlotCdf(self, label=None)
- Draws a Cdf with vertical lines at the observed test stat.
- RunModel(self)
- Run the model of the null hypothesis.
returns: simulated data
- TestStatistic(self, data)
- Computes the test statistic.
data: data in whatever form is relevant
- __init__(self, data)
- Initializes.
data: data in whatever form is relevant
Data descriptors defined here:
- __dict__
- dictionary for instance variables (if defined)
- __weakref__
- list of weak references to the object (if defined)
|
class Interpolator(__builtin__.object) |
| |
Represents a mapping between sorted sequences; performs linear interp.
Attributes:
xs: sorted list
ys: sorted list
|
| |
Methods defined here:
- Lookup(self, x)
- Looks up x and returns the corresponding value of y.
- Reverse(self, y)
- Looks up y and returns the corresponding value of x.
- __init__(self, xs, ys)
Data descriptors defined here:
- __dict__
- dictionary for instance variables (if defined)
- __weakref__
- list of weak references to the object (if defined)
|
class Joint(Pmf) |
| |
Represents a joint distribution.
The values are sequences (usually tuples)
|
| |
- Method resolution order:
- Joint
- Pmf
- _DictWrapper
- __builtin__.object
Methods defined here:
- Conditional(self, i, j, val, label=None)
- Gets the conditional distribution of the indicated variable.
Distribution of vs[i], conditioned on vs[j] = val.
i: index of the variable we want
j: which variable is conditioned on
val: the value the jth variable has to have
Returns: Pmf
- Marginal(self, i, label=None)
- Gets the marginal distribution of the indicated variable.
i: index of the variable we want
Returns: Pmf
- MaxLikeInterval(self, percentage=90)
- Returns the maximum-likelihood credible interval.
If percentage=90, computes a 90% CI containing the values
with the highest likelihoods.
percentage: float between 0 and 100
Returns: list of values from the suite
Methods inherited from Pmf:
- AddConstant(self, other)
- Computes the Pmf of the sum a constant and values from self.
other: a number
returns: new Pmf
- AddPmf(self, other)
- Computes the Pmf of the sum of values drawn from self and other.
other: another Pmf
returns: new Pmf
- CredibleInterval(self, percentage=90)
- Computes the central credible interval.
If percentage=90, computes the 90% CI.
Args:
percentage: float between 0 and 100
Returns:
sequence of two floats, low and high
- DivPmf(self, other)
- Computes the Pmf of the ratio of values drawn from self and other.
other: another Pmf
returns: new Pmf
- Max(self, k)
- Computes the CDF of the maximum of k selections from this dist.
k: int
returns: new Cdf
- MaximumLikelihood(self)
- Returns the value with the highest probability.
Returns: float probability
- Mean(self)
- Computes the mean of a PMF.
Returns:
float mean
- MulConstant(self, other)
- Computes the Pmf of the product of a constant and values from self.
other: a number
returns: new Pmf
- MulPmf(self, other)
- Computes the Pmf of the diff of values drawn from self and other.
other: another Pmf
returns: new Pmf
- Normalize(self, fraction=1.0)
- Normalizes this PMF so the sum of all probs is fraction.
Args:
fraction: what the total should be after normalization
Returns: the total probability before normalizing
- Percentile(self, percentage)
- Computes a percentile of a given Pmf.
Note: this is not super efficient. If you are planning
to compute more than a few percentiles, compute the Cdf.
percentage: float 0-100
returns: value from the Pmf
- Prob(self, x, default=0)
- Gets the probability associated with the value x.
Args:
x: number value
default: value to return if the key is not there
Returns:
float probability
- ProbGreater(self, x)
- Probability that a sample from this Pmf exceeds x.
x: number
returns: float probability
- ProbLess(self, x)
- Probability that a sample from this Pmf is less than x.
x: number
returns: float probability
- Probs(self, xs)
- Gets probabilities for a sequence of values.
- Random(self)
- Chooses a random element from this PMF.
Note: this is not very efficient. If you plan to call
this more than a few times, consider converting to a CDF.
Returns:
float value from the Pmf
- Std(self, mu=None)
- Computes the standard deviation of a PMF.
mu: the point around which the variance is computed;
if omitted, computes the mean
returns: float standard deviation
- SubPmf(self, other)
- Computes the Pmf of the diff of values drawn from self and other.
other: another Pmf
returns: new Pmf
- Var(self, mu=None)
- Computes the variance of a PMF.
mu: the point around which the variance is computed;
if omitted, computes the mean
returns: float variance
- __add__(self, other)
- Computes the Pmf of the sum of values drawn from self and other.
other: another Pmf or a scalar
returns: new Pmf
- __div__(self, other)
- Computes the Pmf of the ratio of values drawn from self and other.
other: another Pmf
returns: new Pmf
- __ge__(self, obj)
- Greater than or equal.
obj: number or _DictWrapper
returns: float probability
- __gt__(self, obj)
- Greater than.
obj: number or _DictWrapper
returns: float probability
- __le__(self, obj)
- Less than or equal.
obj: number or _DictWrapper
returns: float probability
- __lt__(self, obj)
- Less than.
obj: number or _DictWrapper
returns: float probability
- __mul__(self, other)
- Computes the Pmf of the product of values drawn from self and other.
other: another Pmf
returns: new Pmf
- __sub__(self, other)
- Computes the Pmf of the diff of values drawn from self and other.
other: another Pmf
returns: new Pmf
- __truediv__ = __div__(self, other)
- Computes the Pmf of the ratio of values drawn from self and other.
other: another Pmf
returns: new Pmf
Methods inherited from _DictWrapper:
- Copy(self, label=None)
- Returns a copy.
Make a shallow copy of d. If you want a deep copy of d,
use copy.deepcopy on the whole object.
label: string label for the new Hist
returns: new _DictWrapper with the same type
- Exp(self, m=None)
- Exponentiates the probabilities.
m: how much to shift the ps before exponentiating
If m is None, normalizes so that the largest prob is 1.
- GetDict(self)
- Gets the dictionary.
- Incr(self, x, term=1)
- Increments the freq/prob associated with the value x.
Args:
x: number value
term: how much to increment by
- Items(self)
- Gets an unsorted sequence of (value, freq/prob) pairs.
- Largest(self, n=10)
- Returns the largest n values, with frequency/probability.
n: number of items to return
- Log(self, m=None)
- Log transforms the probabilities.
Removes values with probability 0.
Normalizes so that the largest logprob is 0.
- MakeCdf(self, label=None)
- Makes a Cdf.
- MaxLike(self)
- Returns the largest frequency/probability in the map.
- Mult(self, x, factor)
- Scales the freq/prob associated with the value x.
Args:
x: number value
factor: how much to multiply by
- Print(self)
- Prints the values and freqs/probs in ascending order.
- Remove(self, x)
- Removes a value.
Throws an exception if the value is not there.
Args:
x: value to remove
- Render(self, **options)
- Generates a sequence of points suitable for plotting.
Note: options are ignored
Returns:
tuple of (sorted value sequence, freq/prob sequence)
- Scale(self, factor)
- Multiplies the values by a factor.
factor: what to multiply by
Returns: new object
- Set(self, x, y=0)
- Sets the freq/prob associated with the value x.
Args:
x: number value
y: number freq or prob
- SetDict(self, d)
- Sets the dictionary.
- Smallest(self, n=10)
- Returns the smallest n values, with frequency/probability.
n: number of items to return
- Total(self)
- Returns the total of the frequencies/probabilities in the map.
- Values(self)
- Gets an unsorted sequence of values.
Note: one source of confusion is that the keys of this
dictionary are the values of the Hist/Pmf, and the
values of the dictionary are frequencies/probabilities.
- __contains__(self, value)
- __delitem__(self, value)
- __eq__(self, other)
- __getitem__(self, value)
- __hash__(self)
- __init__(self, obj=None, label=None)
- Initializes the distribution.
obj: Hist, Pmf, Cdf, Pdf, dict, pandas Series, list of pairs
label: string label
- __iter__(self)
- __len__(self)
- __repr__ = __str__(self)
- __setitem__(self, value, prob)
- __str__(self)
- iterkeys(self)
- Returns an iterator over keys.
Data descriptors inherited from _DictWrapper:
- __dict__
- dictionary for instance variables (if defined)
- __weakref__
- list of weak references to the object (if defined)
|
class NormalPdf(Pdf) |
| |
Represents the PDF of a Normal distribution.
|
| |
- Method resolution order:
- NormalPdf
- Pdf
- __builtin__.object
Methods defined here:
- Density(self, xs)
- Evaluates this Pdf at xs.
xs: scalar or sequence of floats
returns: float or NumPy array of probability density
- GetLinspace(self)
- Get a linspace for plotting.
Returns: numpy array
- __init__(self, mu=0, sigma=1, label=None)
- Constructs a Normal Pdf with given mu and sigma.
mu: mean
sigma: standard deviation
label: string
- __str__(self)
Methods inherited from Pdf:
- Items(self)
- Generates a sequence of (value, probability) pairs.
- MakePmf(self, **options)
- Makes a discrete version of this Pdf.
options can include
label: string
low: low end of range
high: high end of range
n: number of places to evaluate
Returns: new Pmf
- Render(self, **options)
- Generates a sequence of points suitable for plotting.
If options includes low and high, it must also include n;
in that case the density is evaluated an n locations between
low and high, including both.
If options includes xs, the density is evaluate at those location.
Otherwise, self.GetLinspace is invoked to provide the locations.
Returns:
tuple of (xs, densities)
Data descriptors inherited from Pdf:
- __dict__
- dictionary for instance variables (if defined)
- __weakref__
- list of weak references to the object (if defined)
|
class Pdf(__builtin__.object) |
| |
Represents a probability density function (PDF).
|
| |
Methods defined here:
- Density(self, x)
- Evaluates this Pdf at x.
Returns: float or NumPy array of probability density
- GetLinspace(self)
- Get a linspace for plotting.
Not all subclasses of Pdf implement this.
Returns: numpy array
- Items(self)
- Generates a sequence of (value, probability) pairs.
- MakePmf(self, **options)
- Makes a discrete version of this Pdf.
options can include
label: string
low: low end of range
high: high end of range
n: number of places to evaluate
Returns: new Pmf
- Render(self, **options)
- Generates a sequence of points suitable for plotting.
If options includes low and high, it must also include n;
in that case the density is evaluated an n locations between
low and high, including both.
If options includes xs, the density is evaluate at those location.
Otherwise, self.GetLinspace is invoked to provide the locations.
Returns:
tuple of (xs, densities)
Data descriptors defined here:
- __dict__
- dictionary for instance variables (if defined)
- __weakref__
- list of weak references to the object (if defined)
|
class Pmf(_DictWrapper) |
| |
Represents a probability mass function.
Values can be any hashable type; probabilities are floating-point.
Pmfs are not necessarily normalized.
|
| |
- Method resolution order:
- Pmf
- _DictWrapper
- __builtin__.object
Methods defined here:
- AddConstant(self, other)
- Computes the Pmf of the sum a constant and values from self.
other: a number
returns: new Pmf
- AddPmf(self, other)
- Computes the Pmf of the sum of values drawn from self and other.
other: another Pmf
returns: new Pmf
- CredibleInterval(self, percentage=90)
- Computes the central credible interval.
If percentage=90, computes the 90% CI.
Args:
percentage: float between 0 and 100
Returns:
sequence of two floats, low and high
- DivPmf(self, other)
- Computes the Pmf of the ratio of values drawn from self and other.
other: another Pmf
returns: new Pmf
- Max(self, k)
- Computes the CDF of the maximum of k selections from this dist.
k: int
returns: new Cdf
- MaximumLikelihood(self)
- Returns the value with the highest probability.
Returns: float probability
- Mean(self)
- Computes the mean of a PMF.
Returns:
float mean
- MulConstant(self, other)
- Computes the Pmf of the product of a constant and values from self.
other: a number
returns: new Pmf
- MulPmf(self, other)
- Computes the Pmf of the diff of values drawn from self and other.
other: another Pmf
returns: new Pmf
- Normalize(self, fraction=1.0)
- Normalizes this PMF so the sum of all probs is fraction.
Args:
fraction: what the total should be after normalization
Returns: the total probability before normalizing
- Percentile(self, percentage)
- Computes a percentile of a given Pmf.
Note: this is not super efficient. If you are planning
to compute more than a few percentiles, compute the Cdf.
percentage: float 0-100
returns: value from the Pmf
- Prob(self, x, default=0)
- Gets the probability associated with the value x.
Args:
x: number value
default: value to return if the key is not there
Returns:
float probability
- ProbGreater(self, x)
- Probability that a sample from this Pmf exceeds x.
x: number
returns: float probability
- ProbLess(self, x)
- Probability that a sample from this Pmf is less than x.
x: number
returns: float probability
- Probs(self, xs)
- Gets probabilities for a sequence of values.
- Random(self)
- Chooses a random element from this PMF.
Note: this is not very efficient. If you plan to call
this more than a few times, consider converting to a CDF.
Returns:
float value from the Pmf
- Std(self, mu=None)
- Computes the standard deviation of a PMF.
mu: the point around which the variance is computed;
if omitted, computes the mean
returns: float standard deviation
- SubPmf(self, other)
- Computes the Pmf of the diff of values drawn from self and other.
other: another Pmf
returns: new Pmf
- Var(self, mu=None)
- Computes the variance of a PMF.
mu: the point around which the variance is computed;
if omitted, computes the mean
returns: float variance
- __add__(self, other)
- Computes the Pmf of the sum of values drawn from self and other.
other: another Pmf or a scalar
returns: new Pmf
- __div__(self, other)
- Computes the Pmf of the ratio of values drawn from self and other.
other: another Pmf
returns: new Pmf
- __ge__(self, obj)
- Greater than or equal.
obj: number or _DictWrapper
returns: float probability
- __gt__(self, obj)
- Greater than.
obj: number or _DictWrapper
returns: float probability
- __le__(self, obj)
- Less than or equal.
obj: number or _DictWrapper
returns: float probability
- __lt__(self, obj)
- Less than.
obj: number or _DictWrapper
returns: float probability
- __mul__(self, other)
- Computes the Pmf of the product of values drawn from self and other.
other: another Pmf
returns: new Pmf
- __sub__(self, other)
- Computes the Pmf of the diff of values drawn from self and other.
other: another Pmf
returns: new Pmf
- __truediv__ = __div__(self, other)
Methods inherited from _DictWrapper:
- Copy(self, label=None)
- Returns a copy.
Make a shallow copy of d. If you want a deep copy of d,
use copy.deepcopy on the whole object.
label: string label for the new Hist
returns: new _DictWrapper with the same type
- Exp(self, m=None)
- Exponentiates the probabilities.
m: how much to shift the ps before exponentiating
If m is None, normalizes so that the largest prob is 1.
- GetDict(self)
- Gets the dictionary.
- Incr(self, x, term=1)
- Increments the freq/prob associated with the value x.
Args:
x: number value
term: how much to increment by
- Items(self)
- Gets an unsorted sequence of (value, freq/prob) pairs.
- Largest(self, n=10)
- Returns the largest n values, with frequency/probability.
n: number of items to return
- Log(self, m=None)
- Log transforms the probabilities.
Removes values with probability 0.
Normalizes so that the largest logprob is 0.
- MakeCdf(self, label=None)
- Makes a Cdf.
- MaxLike(self)
- Returns the largest frequency/probability in the map.
- Mult(self, x, factor)
- Scales the freq/prob associated with the value x.
Args:
x: number value
factor: how much to multiply by
- Print(self)
- Prints the values and freqs/probs in ascending order.
- Remove(self, x)
- Removes a value.
Throws an exception if the value is not there.
Args:
x: value to remove
- Render(self, **options)
- Generates a sequence of points suitable for plotting.
Note: options are ignored
Returns:
tuple of (sorted value sequence, freq/prob sequence)
- Scale(self, factor)
- Multiplies the values by a factor.
factor: what to multiply by
Returns: new object
- Set(self, x, y=0)
- Sets the freq/prob associated with the value x.
Args:
x: number value
y: number freq or prob
- SetDict(self, d)
- Sets the dictionary.
- Smallest(self, n=10)
- Returns the smallest n values, with frequency/probability.
n: number of items to return
- Total(self)
- Returns the total of the frequencies/probabilities in the map.
- Values(self)
- Gets an unsorted sequence of values.
Note: one source of confusion is that the keys of this
dictionary are the values of the Hist/Pmf, and the
values of the dictionary are frequencies/probabilities.
- __contains__(self, value)
- __delitem__(self, value)
- __eq__(self, other)
- __getitem__(self, value)
- __hash__(self)
- __init__(self, obj=None, label=None)
- Initializes the distribution.
obj: Hist, Pmf, Cdf, Pdf, dict, pandas Series, list of pairs
label: string label
- __iter__(self)
- __len__(self)
- __repr__ = __str__(self)
- __setitem__(self, value, prob)
- __str__(self)
- iterkeys(self)
- Returns an iterator over keys.
Data descriptors inherited from _DictWrapper:
- __dict__
- dictionary for instance variables (if defined)
- __weakref__
- list of weak references to the object (if defined)
|
class Suite(Pmf) |
| |
Represents a suite of hypotheses and their probabilities.
|
| |
- Method resolution order:
- Suite
- Pmf
- _DictWrapper
- __builtin__.object
Methods defined here:
- Likelihood(self, data, hypo)
- Computes the likelihood of the data under the hypothesis.
hypo: some representation of the hypothesis
data: some representation of the data
- LogLikelihood(self, data, hypo)
- Computes the log likelihood of the data under the hypothesis.
hypo: some representation of the hypothesis
data: some representation of the data
- LogUpdate(self, data)
- Updates a suite of hypotheses based on new data.
Modifies the suite directly; if you want to keep the original, make
a copy.
Note: unlike Update, LogUpdate does not normalize.
Args:
data: any representation of the data
- LogUpdateSet(self, dataset)
- Updates each hypothesis based on the dataset.
Modifies the suite directly; if you want to keep the original, make
a copy.
dataset: a sequence of data
returns: None
- MakeOdds(self)
- Transforms from probabilities to odds.
Values with prob=0 are removed.
- MakeProbs(self)
- Transforms from odds to probabilities.
- Print(self)
- Prints the hypotheses and their probabilities.
- Update(self, data)
- Updates each hypothesis based on the data.
data: any representation of the data
returns: the normalizing constant
- UpdateSet(self, dataset)
- Updates each hypothesis based on the dataset.
This is more efficient than calling Update repeatedly because
it waits until the end to Normalize.
Modifies the suite directly; if you want to keep the original, make
a copy.
dataset: a sequence of data
returns: the normalizing constant
Methods inherited from Pmf:
- AddConstant(self, other)
- Computes the Pmf of the sum a constant and values from self.
other: a number
returns: new Pmf
- AddPmf(self, other)
- Computes the Pmf of the sum of values drawn from self and other.
other: another Pmf
returns: new Pmf
- CredibleInterval(self, percentage=90)
- Computes the central credible interval.
If percentage=90, computes the 90% CI.
Args:
percentage: float between 0 and 100
Returns:
sequence of two floats, low and high
- DivPmf(self, other)
- Computes the Pmf of the ratio of values drawn from self and other.
other: another Pmf
returns: new Pmf
- Max(self, k)
- Computes the CDF of the maximum of k selections from this dist.
k: int
returns: new Cdf
- MaximumLikelihood(self)
- Returns the value with the highest probability.
Returns: float probability
- Mean(self)
- Computes the mean of a PMF.
Returns:
float mean
- MulConstant(self, other)
- Computes the Pmf of the product of a constant and values from self.
other: a number
returns: new Pmf
- MulPmf(self, other)
- Computes the Pmf of the diff of values drawn from self and other.
other: another Pmf
returns: new Pmf
- Normalize(self, fraction=1.0)
- Normalizes this PMF so the sum of all probs is fraction.
Args:
fraction: what the total should be after normalization
Returns: the total probability before normalizing
- Percentile(self, percentage)
- Computes a percentile of a given Pmf.
Note: this is not super efficient. If you are planning
to compute more than a few percentiles, compute the Cdf.
percentage: float 0-100
returns: value from the Pmf
- Prob(self, x, default=0)
- Gets the probability associated with the value x.
Args:
x: number value
default: value to return if the key is not there
Returns:
float probability
- ProbGreater(self, x)
- Probability that a sample from this Pmf exceeds x.
x: number
returns: float probability
- ProbLess(self, x)
- Probability that a sample from this Pmf is less than x.
x: number
returns: float probability
- Probs(self, xs)
- Gets probabilities for a sequence of values.
- Random(self)
- Chooses a random element from this PMF.
Note: this is not very efficient. If you plan to call
this more than a few times, consider converting to a CDF.
Returns:
float value from the Pmf
- Std(self, mu=None)
- Computes the standard deviation of a PMF.
mu: the point around which the variance is computed;
if omitted, computes the mean
returns: float standard deviation
- SubPmf(self, other)
- Computes the Pmf of the diff of values drawn from self and other.
other: another Pmf
returns: new Pmf
- Var(self, mu=None)
- Computes the variance of a PMF.
mu: the point around which the variance is computed;
if omitted, computes the mean
returns: float variance
- __add__(self, other)
- Computes the Pmf of the sum of values drawn from self and other.
other: another Pmf or a scalar
returns: new Pmf
- __div__(self, other)
- Computes the Pmf of the ratio of values drawn from self and other.
other: another Pmf
returns: new Pmf
- __ge__(self, obj)
- Greater than or equal.
obj: number or _DictWrapper
returns: float probability
- __gt__(self, obj)
- Greater than.
obj: number or _DictWrapper
returns: float probability
- __le__(self, obj)
- Less than or equal.
obj: number or _DictWrapper
returns: float probability
- __lt__(self, obj)
- Less than.
obj: number or _DictWrapper
returns: float probability
- __mul__(self, other)
- Computes the Pmf of the product of values drawn from self and other.
other: another Pmf
returns: new Pmf
- __sub__(self, other)
- Computes the Pmf of the diff of values drawn from self and other.
other: another Pmf
returns: new Pmf
- __truediv__ = __div__(self, other)
- Computes the Pmf of the ratio of values drawn from self and other.
other: another Pmf
returns: new Pmf
Methods inherited from _DictWrapper:
- Copy(self, label=None)
- Returns a copy.
Make a shallow copy of d. If you want a deep copy of d,
use copy.deepcopy on the whole object.
label: string label for the new Hist
returns: new _DictWrapper with the same type
- Exp(self, m=None)
- Exponentiates the probabilities.
m: how much to shift the ps before exponentiating
If m is None, normalizes so that the largest prob is 1.
- GetDict(self)
- Gets the dictionary.
- Incr(self, x, term=1)
- Increments the freq/prob associated with the value x.
Args:
x: number value
term: how much to increment by
- Items(self)
- Gets an unsorted sequence of (value, freq/prob) pairs.
- Largest(self, n=10)
- Returns the largest n values, with frequency/probability.
n: number of items to return
- Log(self, m=None)
- Log transforms the probabilities.
Removes values with probability 0.
Normalizes so that the largest logprob is 0.
- MakeCdf(self, label=None)
- Makes a Cdf.
- MaxLike(self)
- Returns the largest frequency/probability in the map.
- Mult(self, x, factor)
- Scales the freq/prob associated with the value x.
Args:
x: number value
factor: how much to multiply by
- Remove(self, x)
- Removes a value.
Throws an exception if the value is not there.
Args:
x: value to remove
- Render(self, **options)
- Generates a sequence of points suitable for plotting.
Note: options are ignored
Returns:
tuple of (sorted value sequence, freq/prob sequence)
- Scale(self, factor)
- Multiplies the values by a factor.
factor: what to multiply by
Returns: new object
- Set(self, x, y=0)
- Sets the freq/prob associated with the value x.
Args:
x: number value
y: number freq or prob
- SetDict(self, d)
- Sets the dictionary.
- Smallest(self, n=10)
- Returns the smallest n values, with frequency/probability.
n: number of items to return
- Total(self)
- Returns the total of the frequencies/probabilities in the map.
- Values(self)
- Gets an unsorted sequence of values.
Note: one source of confusion is that the keys of this
dictionary are the values of the Hist/Pmf, and the
values of the dictionary are frequencies/probabilities.
- __contains__(self, value)
- __delitem__(self, value)
- __eq__(self, other)
- __getitem__(self, value)
- __hash__(self)
- __init__(self, obj=None, label=None)
- Initializes the distribution.
obj: Hist, Pmf, Cdf, Pdf, dict, pandas Series, list of pairs
label: string label
- __iter__(self)
- __len__(self)
- __repr__ = __str__(self)
- __setitem__(self, value, prob)
- __str__(self)
- iterkeys(self)
- Returns an iterator over keys.
Data descriptors inherited from _DictWrapper:
- __dict__
- dictionary for instance variables (if defined)
- __weakref__
- list of weak references to the object (if defined)
|
|