Calculate Transmission Polynomial

[Gesner Passos]

In calculate transmission, the polynomial order is to be applied to the log of the data, as specified by the scientist:

Are we actually fitting the polynomial to the Log(Trans) ? - Which is what I intended.

[Gesner Passos]

re #8977: Make Polynomial of Log(trans) for calculate transmission

Changeset: 1d1acca699d159cb604a4a01a5c19f0f814a0ab7

[Gesner Passos]

While doing this ticket I've spotted an issue with the error propagation #9011

[Gesner Passos]

Tester:

If rebining is not present, fitting the CalculateTransmission can be reproduced outside as:

• Log(xxx_unfitted)
• Fit
• Exp

The check_calculate_transmission shows how I did this to check the result. The result should be the same except for the error because of the issue presented in #9011.

[Gesner Passos]

It broke the systemtest LOQReductionGUI.LOQMinimalBatchReduction.

It is ok. That specific test uses as Mask the MaskLOQData.txt which has the following line:

FIT/TRANS/CAN/POLYNOMIAL4 2.2 10.0

Now, we have changed the polynomial to be applied in the Log(trans). Hence, having the results changing is a good signal.

[Gesner Passos]

re #8977: Update file that uses polynomial fit.

Changeset: f4bc5d79fd1c4a632c33a936c066198b9a88a125

[Gesner Passos]

#9030 will cause the Errors in Fit to change, this will effect CalculateTransmission. Hence, before applying this ticket, specially because it changes Systemtest data, it is better to wait for #9030.

[Gesner Passos]

Tester:

You can use check8977.py to check this ticket as well, it is a variation of the one proposed in #9011 to allow checking for the Polynomial fit.

[Gesner Passos]

Peter:

Because #9030 is delayed (waiting for the green signal from Martyn), I won't be able to push this one. comment:2 is the commit that implements what is required in this ticket. You will probably need to run again the systemtest that produced the comment:9 because it was done before the addition of #9030.

As soon as this pass, please inform Richard who is keen to test this new feature.

[Gesner Passos]

The check8977_real_example.py allows you to check against a real use-case proposed from the user.

[Gesner Passos]

What we do not want as answer. But it depends on #9030.

[Peter Parker]

I have yet to look into this, but it seems that Gesner had possibly completed this but was just waiting for another ticket to clear. I will take a look after the release.

[Nick Draper]

Moved to the backlog at the code freeze of R3.3

[Peter Parker]

Merge branch 'master' into feature/8977_calculate_transmission_poly

Conflicts:

Code/Mantid/Framework/Algorithms/src/CalculateTransmission.cpp

Refs #8977

Changeset: 3ccb3b011120bb73360e721a6951ec7e213fafe6

[Stuart Campbell]

This ticket has been transferred to github issue ​9820