UtilPlot Module

poisson_approval.utils.UtilPlot.plt_cdf(data, weights, n_samples, data_min=0, data_max=1, **kwargs)[source]

Plot a cumulative distribution function from Monte-Carlo experiments, with error area.

We assume that the values in data are obtained by a Monte-Carlo method on a random variable that is in [data_min, data_max]. For the confidence interval, cf. plt_step_with_error().

Parameters

  • data (array-like) – Each encountered value.

  • weights (array-like) – The weight (probability) with which the value was encountered.

  • n_samples (int) – Number of samples used for Monte-Carlo.

  • data_min (Number) – Minimum possible value of data. Default: 0.

  • data_max (Number) – Maximum possible value of data. Default: 1.

  • kwargs – Other keyword arguments are passed to the function step of matplotlib.

Examples

>>> data = np.random.random(100)
>>> weights = np.ones(100) / 100
>>> plt_cdf(data, weights, n_samples=100)
poisson_approval.utils.UtilPlot.plt_plot_with_error(x, y, n_samples, **kwargs)[source]

Adaptation of plt.plot for Monte-Carlo experiments, with error area.

We assume that y-values are obtained by a Monte-Carlo method on a random variable that is in [0, 1]. The margin of error is then bounded by the one of a repeated Bernouilli of parameter 0.5, which is itself approximated by a normal distribution. Hence the 95% confidence interval is approximately ± 1 / sqrt(n_samples). Note that this confidence interval is also valid (only overestimated) when the y-value is 0 or 1.

Parameters

  • x (array-like) – The x-data.

  • y (array-like) – The y-data. Values must be in [0, 1].

  • n_samples (int) – Number of samples used for Monte-Carlo.

  • kwargs – Other keyword arguments are passed to the function plot of matplotlib.

Examples

>>> x = np.arange(0, 1, .1)
>>> y = np.random.random(10)
>>> plt_plot_with_error(x, y, n_samples=100)
poisson_approval.utils.UtilPlot.plt_step_with_error(x, y, n_samples, **kwargs)[source]

Adaptation of plt.step for Monte-Carlo experiments, with error area.

We assume that y-values are obtained by a Monte-Carlo method on a random variable that is in [0, 1]. The margin of error is then bounded by the one of a repeated Bernouilli of parameter 0.5, which is itself approximated by a normal distribution. Hence the 95% confidence interval is approximately ± 1 / sqrt(n_samples). Note that this confidence interval is also valid (only overestimated) when the y-value is 0 or 1.

Parameters

  • x (array-like) – The x-data.

  • y (array-like) – The y-data. Values must be in [0, 1].

  • n_samples (int) – Number of samples used for Monte-Carlo.

  • kwargs – Other keyword arguments are passed to the function step of matplotlib.

Examples

>>> x = np.arange(0, 1, .1)
>>> y = np.random.random(10)
>>> plt_step_with_error(x, y, n_samples=100)