When calculating the rate, right now it assumes that M0-5 is in the exponential part only. Need to make it more flexible so that it can calculate if M0-5 is in the characteristic part at all (Mmeanchar 5.15 or something).

Travis CI stops the build because no output has been received for 10 minutes:

Running HAZ on: ../tests/Set3/S3Test3/Approach a/Run_S3Test3.txt
At line 526 of file /home/travis/build/arkottke/HAZ/HAZ files/readFlt.f (unit = 10, file = 'Source_S3Test3.txt')
Fortran runtime error: End of file
Running HAZ on: ../tests/Set3/S3Test3/Approach b1/Run_S3Test3.txt
Running HAZ on: ../tests/Set3/S3Test1a/Run_S3Test1a.txt
Running HAZ on: ../tests/Set3/S3Test1b/Run_S3Test1b.txt
Running HAZ on: ../tests/Set3/S3Test2/Fractiles/Run_Fractiles.txt
STOP 99
Running HAZ on: ../tests/Set3/S3Test2/Hazard/Run_S3Test2.txt
No output has been received in the last 10m0s, this potentially indicates a stalled build or something wrong with the build itself.
Check the details on how to adjust your build configuration on: https://docs.travis-ci.com/user/common-build-problems/#Build-times-out-because-no-output-was-received

Revisit this line of code, make sure robust for fault that are not rectangles

1. testing source type 7 (Nick)
2. testing source type 3 (Christie)
3. testing WAACY mag pdf (Debra)
4. testing NGA-W1 GMPEs (Christie)
5. bilinear TE mag pdf (Debra)
6. testing epistemic uncer in median and sigma (Christie)
7. testing segmentation for faults (Christie)
8. testing all four rate methods for faults (SR, Act rate, Rec Int, Moment rate) (Debra)
9. Run Option 1 (Christie)
10. Run Option 2 (Christie)

Some of the calls to this subroutine in MED_ATT.F have five arguments while the subroutine itself in SIGMA.f only has three arguments. Need to clean up sigma models, run verification tests, and add identifying numbers to jcalc list.

Check for limit on Ztor (20 km)

Incorporate fixes from Nick's email, which are causing problems on other compilers

In many instances, arrays are initialized using a single length (e.g., wa_pmag(1)), but the actual length is known and not 1. This manner of programming prevents the compiler from performing optimizations and should be addressed.

There is an issue with the deaggregation bins (out4) in that the gm_wt is not accounted for in the variable HazBins. The likely fix is to modify the file HAZ_MAIN2.f line 543 to add dble(mHaz wt gm_wt(iProb,jType,iAtten)) to HazBins. This issue does not affect the average deaggregation values (Mbar, Dbar, etc.), and was not caught in verification test 2.1 because that test only had one ground motion model. Also noticed that the subroutine output_HazBins passes in the variable jcalc but this is never used - should clean this up.

The override mag option does not work (and no one has used this for years)
Remove this option and remove the input in the fault file

1. Test more fault orientations (make sure to test ones with straight horizontal and vertical segments)
2. Need to make robust so that if only 1 grid cell down-dip or along strike it still works, or at least gives an error

In Run_S1Test1.txt, the magnitude deaggregation values are defined as 4.75, 5.75, 6.25, 6.75, and 7.0. The Set1Test1_Site1.out4 file lists these magnitude ranges to one decimal place as 4.8, 5.8, 6.3, 6.8, and 7.0. These input and output values should be consistent. Additionally, when the code is compiled and run using gfortran via msys2 the Set1Test1_Site1.out4 file lists these magnitude ranges as 4.8, 5.8, 6.2, 6.8, 7.0 (code = develop branch Feb 9). This is likely a compiler compatibility issue but it should be checked that it is not a change to the code. Possible fix is to change the formatting such that two decimal places are written to the output file.

The HAZ45.2 executables that were distributed on box were compiled with array conformance options. Ref Test 9 will not run on these executables as the code will produce an array conformance error. The error is coming from the variable period1(jType,iProb) at the very end of the meanInten subroutine. The issue is that iProb is not needed with run option 1, never set, and is therefore given a value of zero. A zero index is not a valid index. If the code is compiled without array conformance options then there is no issue, but the code should be fixed to run with array conformance options. Possible fix is to set iProb = 1 in the determ subroutine for Run Option 1.

Think about how to rework. grid_top should be based on iflt and j. Second array identifier shouldn't always be 1.

in SetFltBottom subroutine use the haversine formula instead of flat earth

Hypodepth error fixed for SourceType=1 - need to fix other SourceTypes

Add multiple dips to areal zone corresponding to NML, SS, RV
This will allow the dip to be properly correlated with the mech if there are multiple mech used

delete sourcetype 6 and create an option to run non-uniform weighting of rupture locations with sourcetype 5

Test cases have out of data version string.

Make the mixture model weights more flexible (inputs)