search

imr-framework / pypulseq

Pulseq in Python
https://pypulseq.readthedocs.io
GNU Affero General Public License v3.0
118 stars 63 forks source link

'split_gradient_at' function does not work well #25

Closed opennog closed 4 years ago

opennog commented 4 years ago

Describe the bug split_gradient_at function does not split the gradient defined as 'trap' correctly.

Expected behavior For example, if the gradient maximum value is gy.amplitude=138888 and the rise and fall times are of 3xraster times, when you split in two equal parts the values of, ie., the fall part are array([115740.74074074, 69444.44444444, 23148.14814815]). See that the maximum values is not 138888 anymore, and that it does not end in zero either.

sravan953 commented 4 years ago

Could you please share code to reproduce this on pypulseq v1.2.0?

opennog commented 4 years ago

For example: system = Opts(max_grad=32, grad_unit='mT/m', max_slew=150, slew_unit='T/m/s', rf_ringdown_time=20e-6, rf_dead_time=100e-6, adc_dead_time=10e-6, rf_raster_time=1e-6)

gy = make_trapezoid(channel='y', system=system, area=4.166666666666667)

gy_parts = split_gradient_at(grad=gy, time_point=6e-05 / 2, system=system)

Here, gy is: namespace(amplitude=138888.88888888888, area=4.166666666666667, channel='y', delay=0.0, fall_time=3.0000000000000004e-05, first=0, flat_area=0.0, flat_time=0.0, last=0, rise_time=3.0000000000000004e-05, type='trap')

and gy_parts is: (namespace(channel='y', delay=0.0005300000000000001, first=0.0, last=138888.88888888888, t=array([0.e+00, 1.e-05, 2.e-05]), type='grad', waveform=array([ 23148.14814815, 69444.44444444, 115740.74074074])), namespace(channel='y', delay=0, first=138888.88888888888, last=0.0, t=array([0.e+00, 1.e-05, 2.e-05]), type='grad', waveform=array([115740.74074074, 69444.44444444, 23148.14814815])))

As you can see, gy max is 138888.88888 then, even gy_parts says its first and last values (on the right and left parts, respectively) are of 138888.888888. However, the actual values of the arrays ([ 23148.14814815, 69444.44444444, 115740.74074074] and [115740.74074074, 69444.44444444, 23148.14814815]) are scaled differently. Probably because it changes the shape of the triangle.

As far as I understand it, this way we are loosing one raster time to construct the triangle with a desired area.

Thanks! Óscar

sravan953 commented 4 years ago

Hello, can you please pull from commit 7ee685224962cfa5c500c6976a624fb6d09ff396 and let me know if the issue is fixed? You can download the ZIP from: https://github.com/imr-framework/pypulseq/tree/7ee685224962cfa5c500c6976a624fb6d09ff396

sravan953 commented 4 years ago

Please reopen this issue if the bug is not fixed.