nstat.ConfidenceInterval

class nstat.ConfidenceInterval(time, bounds, *args, color=None, value=0.95)[source]

Bases: object

Confidence interval for a time series or Covariate.

Stores a pair of (lower, upper) bound traces over a shared time axis, with plotting support for both line and shaded-patch styles.

Parameters:

time (array_like) – Time vector.
bounds (array_like, shape (n, 2)) – Lower and upper bounds at each time point.
*args – Positional metadata: (name, xlabelval, xunits, yunits, dataLabels, plotProps). If a single short string is passed it is interpreted as the colour (Matlab compatibility).
color (str or None) – Line / patch colour. Default 'b'.
value (float, default 0.95) – Confidence level (e.g. 0.95 for 95 %).

__init__(time, bounds, *args, color=None, value=0.95)[source]

Parameters:

color (str | None)
value (float)

Return type:

None

property lower: ndarray

property upper: ndarray

setColor(color)[source]

Set the plot colour.

Parameters:: color (str)
Return type:: None

setValue(value)[source]

Set the confidence level (e.g. 0.95 for 95 %).

Parameters:: value (float)
Return type:: None

dataToMatrix()[source]

Return the bounds as an (n, 2) numpy array.

Return type:: ndarray

dataToStructure()[source]

Serialise to a plain dictionary (matches Matlab dataToStructure).

Return type:: dict

getSigInTimeWindow(wMin, wMax, holdVals=0)[source]

Return a new ConfidenceInterval restricted to [wMin, wMax].

Parameters:

wMin (float)
wMax (float)
holdVals (int)

windowedSignal(windowTimes)[source]: Return a ConfidenceInterval restricted to the given time window.

shift(deltaT, updateLabels=False)[source]

Return a time-shifted copy.

Parameters:

deltaT (float)
updateLabels (bool)

resample(sample_rate)[source]

Resample the CI bounds to a new sample rate via linear interpolation.

Parameters:: sample_rate (float)

property derivative: Numerical derivative of both bounds.

merge(other, holdVals=0)[source]

Merge with another ConfidenceInterval along the time axis.

Parameters:: holdVals (int)

copySignal()[source]: Return a deep copy.

getSubSignal(identifier)[source]: Return a single column (lower=0 or upper=1) as a 1-column CI.

setMinTime(minTime=None, holdVals=0)[source]

Restrict to times >= minTime.

Parameters:: holdVals (int)

setMaxTime(maxTime=None, holdVals=0)[source]

Restrict to times <= maxTime.

Parameters:: holdVals (int)

toStructure()[source]

Alias for dataToStructure().

Return type:: dict

plot(color=None, alphaVal=0.2, drawPatches=0, ax=None)[source]

Plot the confidence interval.

Parameters:

color (str or None) – Override colour (default: self.color).
alphaVal (float, default 0.2) – Transparency for shaded patches.
drawPatches (int, default 0) – If 1, draw a shaded fill_between region instead of lines.
ax (Axes or None) – Matplotlib axes. If None, uses plt.gca().

Returns:

Plot handles.

Return type:

PolyCollection or list of Line2D

static fromStructure(structure)[source]

Reconstruct a ConfidenceInterval from a dictionary.

Parameters:: structure (dict)
Return type:: ConfidenceInterval