utils#

py_fatigue.cycle_count.utils.solve_lffd(x: Any) → Any | CycleCount#

Solve the low-frequency fatigue dynamics of a cycle count or return the object as is.

Parameters:

x (Any) – The object to evaluate. If it is a CycleCount instance, the low-frequency fatigue dynamics is solved. Otherwise, the object is returned as is.

Returns:

The object evaluated

Return type:

Any

py_fatigue.cycle_count.utils.aggregate_cc(df: pd.DataFrame, aggr_by: str, save_residuals: bool = False) → pd.DataFrame | tuple[pd.DataFrame, DefaultDict[str, DefaultDict[str, list[float]]]]#

Aggregate a pandas dataframe by time window. The pandas dataframe must have a DatetimeIndex and at least one column whose name starts with ‘CC_’ containing CycleCount instances, e.g.,

Example dataframe to be aggregated#

DateTimeIndex

CC_CycleCount-name-1

CC_CycleCount-name-2

CC_CycleCount-name-3

CC_…

2018-01-01 00:00:00

CycleCount-name-1 (01 Jan 2018, 00:00)

CycleCount-name-2 (01 Jan 2018, 00:00)

CycleCount-name-3 (01 Jan 2018, 00:00)

…

2018-01-01 01:00:00

CycleCount-name-1 (01 Jan 2018, 01:00)

CycleCount-name-2 (01 Jan 2018, 01:00)

CycleCount-name-3 (01 Jan 2018, 01:00)

…

2018-01-01 02:00:00

CycleCount-name-1 (01 Jan 2018, 02:00)

CycleCount-name-2 (01 Jan 2018, 02:00)

CycleCount-name-3 (01 Jan 2018, 02:00)

…

⋮

⋮

⋮

⋮

⋱

The function performs the following workflow:

  1. Perform initial checks on the input pandas dataframe

  2. Build the aggregation dictionary

  3. Aggregate the dataframe by time window, i.e. the aggregated CycleCounts

  4. Retrieve the low-frequency fatigue dynamics on the aggregated dataframe

  5. Save the residuals sequences of each aggregated CycleCount

Parameters:

  • df (pd.DataFrame) – The dataframe to cluster

  • aggr_by (str) – The time window to cluster the dataframe by. It must be a valid pandas date offset frequency string or ‘all’. For all the frequency string aliases offered by pandas, see: pandas-timeseries.html#dateoffset-objects.

  • save_residuals (bool, optional) – If True, the residuals sequences of each aggregated CycleCount are saved

Returns:

The aggregated dataframe and the residuals sequences of each aggregated

Return type:

tuple[pd.DataFrame, dict[str, dict[str, list]]]

py_fatigue.cycle_count.utils.plot_aggregated_residuals(res_dct: dict[str, DefaultDict[str, DefaultDict[str, list[float]]]], plt_prmtr: str, minor_grid: bool = True) → tuple#

Plot the aggregated residuals sequences. T

Parameters:

  • res_dct (dict[str, DefaultDict[str, DefaultDict[str, list[float]]]]) – The residuals sequences of each aggregated CycleCount as returned by aggregate_cc().

  • plt_prmtr (str) – The parameter to plot.

  • labels (Collection[str]) – The labels of the aggregated CycleCounts.

  • minor_grid (bool, optional) – Whether to plot the minor grid, by default True

Returns:

Return type:

tuple[plt.figure.Figure, plt.axes.Axes]

py_fatigue.cycle_count.utils.calc_aggregated_damage(df: pd.DataFrame, sn: dict[str, SNCurve] | DefaultDict[str, SNCurve] | list[SNCurve] | SNCurve) → pd.DataFrame#

Calculate the damage of each aggregated CycleCount. This function needs the output of aggregate_cc().

Parameters:

  • df (pd.DataFrame) – The aggregated CycleCount as returned by aggregate_cc().

  • sn (DefaultDict[str, SNCurve]) – The S-N curves of each aggregated CycleCount.

Returns:

The damage of each aggregated CycleCount as a multi-indexed dataframe.

Return type:

pd.DataFrame