add seasonality with monthly values to example scenario

[acavelan]

The current example scenario uses Fourier coefficients for seasonality. In practice, users may prefer monthly values. This PR adds an example using monthly values.

When using monthly values, the seasonality is shifter to the right by 15 days because the model is assuming the provided monthly values are averages for each month. This PR removes the shift to align with using Fourier coefficients so that both methods produce near identical InputEIR and SimulatedEIR.

Example of XML configuration:

Fourier coefficients

                    <fourierSeries EIRRotateAngle="0">
                        <coeffic a="-1.0313715140024822" b="0.5474993785222372"/>
                        <coeffic a="-0.005155722300211589" b="-0.19769509633382162"/>
                    </fourierSeries>

Monthly values

                    <monthlyValues smoothing="fourier" EIRRotateAngle="0">
                        <value>98</value>
                        <value>132</value>
                        <value>223</value>
                        <value>403</value>
                        <value>1156</value>
                        <value>953</value>
                        <value>755</value>
                        <value>403</value>
                        <value>293</value>
                        <value>149</value>
                        <value>116</value>
                        <value>107</value>
                    </monthlyValues>

Note that EIRRotateAngle becomes available for monthly values (it was previously only available for Fourier coefficients).
In addition, EIRRotateAngle is now optional for both methods and defaults to 0.

Tests that were using monthly values have been updated.

[b-s-code]

I support the inclusion of example monthly values.

The monthly values added here give different output to what is obtained by specifying Fourier coefficients.

The difference isn't huge and I assume it is to be expected? E.g. these output plots look similar but there are some observable differences. Top row = Fourier coefficients in XML, bottom row = monthly values in XML.

I don't suppose there's any reason this might be an issue? None come to my mind.

If no concerns on your end then I'm happy for this to be merged.

[acavelan]

Apparently when using monthly values, InputEIR is shifted slightly. I'm working on a fix.

[acavelan]

openmalaria/model/Transmission/transmission.h

Line 188 in e449fef

EIRRotateAngle = M_PI * (1.0 / 12.0 - 1.0 / 365.0);

When using monthly values, the InputEIR vector is shifted by approximately half a month: this is most likely assuming the monthly values provided by the user are averages for each month.

[b-s-code]

Nice one. Just a print statement left over.

[b-s-code]

Can delete?

[b-s-code]

Looks good.