## Computing the damage

There is a few similar methods for the damage calculation. *(note: damage is reversed value of Number of cycles; D = 1/N)*

They differ only in the load input format. Load can be defined by load1 and load2 (`compute_D_l1_l2()`

), or by apmlitude and R-ratio (`compute_D_amp_R()`

) or by the amplitude and meanstress (`compute_D_amp_mean()`

).

Let's show one of them, e.g. `compute_D_l1_l2()`

:

```
compute_D_l1_l2(self, load, verbose=False, excel_out=None):
Parameters
----------
load : np.ndarray
0 axis ... loadcycles
1 axis ... sequence of [load1, load2, required_cycles, shift_factor]
[[load1, load2, required_cycles, shift_factor], ...]
verbose : bool
if True, prints recalculated SN-curve amplitudes to R=-1 (used for calculation),
default: False
excel_out : string, optional
name of the xlsx output file where calculation details will be written
Returns
-------
df : pd.DataFrame
includes results and some other interesting information
```

You can see that the load is defined in the numpy array. First axis contains separate loadings to be computed, second axis contains an iterable of load1, load2, required_cycles & shift_factor. *(other methods like compute_D_amp_R() contain amplitude and R-ratio instead of load1 and load2)*

There is a verbose option, that warns you in the case that SN curve is given for different R-ratio than R=-1. In that case SN-curve parameters are recalcualted automatically to R=-1 as well as the load.

### output to the excel document

Another interesting option is the possibility to export results into an excel file. Calculation results, used SN-curve parameters for original R and R=-1 and also SN-curve points (e.g. for a fast plot) will be exported in excel document if you pass its name as a parameter *excel_out*.