Lavinia, I implemented your suggestion here:
https://github.com/pyqtgraph/pyqtgraph/pull/1291
If you get a chance, could you see if that does what you expect and comment
over there if you see any issues?
On Thursday, June 4, 2020 at 12:49:34 AM UTC-7, Lavinia B wrote:
>
> Hi Megan,
> Sorry to revive this but I found some issues. While the fix works for the
> example case, in my application I am switching the imageview type within
> the same widget, by switching data passed. The current fix breaks the
> "timeline" version. So axes must be specified as a parameter.
> here is my current take:
>
> axes = (self.axes['x'], self.axes['y'])
> is_row_major = pg.getConfigOption('imageAxisOrder') == 'row-major'
> data = None
> coords = None
> if (is_row_major):
> data, coords = self.roi.getArrayRegion(image.view(np.ndarray),
> self.imageItem,
> (self.axes['y'], self.axes['x']),
> returnMappedCoords=True)
> else:
> data, coords = self.roi.getArrayRegion(image.view(np.ndarray),
> self.imageItem, axes,
> returnMappedCoords=True)
>
>
>
> Also the region of interest had a bug- was sampling wrong for the
> row-major data ( it was sampling from a column major version of the data)-
> so my plots were wrong. In the comments of the getArrayRegion() it is
> mentioned axes have to be passed swapped, so that's what i did and it seems
> fine now.
> Not sure if i should have commented this on your pull request :D But i
> hope you see it!
>
>
> On Tuesday, May 19, 2020 at 11:48:47 PM UTC+2, megan wrote:
>>
>> Hi Lavinia,
>>
>> I think I made a fix and am submitting a pull request to the official
>> pyqtgraph repository. If you want the fix sooner, you can get it from my
>> github:
>>
>> https://github.com/meganbkratz/pyqtgraph/
>> <https://github.com/meganbkratz/pyqtgraph/tree/imageAxisOrder_bugfix>
>>
>> The branch with the bugfix is called "imageAxisOrder_bugfix".
>>
>> Megan
>>
>> On Tue, May 19, 2020 at 1:47 PM megan kratz <[email protected]> wrote:
>>
>>> Hi,
>>>
>>> Thanks for the example! This is definitely a bug. In addition to the
>>> error you're getting about the axes being the wrong shape, there also seems
>>> to be a problem with the wrong data being pulled from the image when
>>> imageAxisOrder='row-major'.
>>>
>>> For now I'd recommend not using the 'row-major' option if that's at all
>>> possible. I'll try to look into it more and figure out a fix.
>>>
>>> Megan
>>>
>>>
>>> On Tue, May 19, 2020 at 11:16 AM Lavinia B <[email protected]> wrote:
>>>
>>>> Hi,
>>>> I managed to make an example of my issue . For some reason I cannot
>>>> get the range to adjust for each color, but I put a border around the
>>>> image, to make it clear where it ends.
>>>> I fill in arrays representing rgba, and I stack that to form my data.
>>>>
>>>> If you set the configuration to row major, rescale the roi to an uneven
>>>> rectangle, then you should get my issue. You can also see in the pictures,
>>>> because of the error the plots stop updating.
>>>>
>>>> I expected it to work the same as in column major. ( you can move
>>>> around the roi, and when the roi overlaps an area outside the image, the
>>>> plot data corresponding to it is 0)
>>>> Any thoughts?
>>>>
>>>>
>>>> On Tuesday, May 19, 2020 at 5:57:41 PM UTC+2, megan wrote:
>>>>>
>>>>> Okay, I've not been able to replicate this. Can you post some example
>>>>> code that produces this error?
>>>>>
>>>>> On Tue, May 19, 2020 at 9:43 AM Lavinia B <[email protected]> wrote:
>>>>>
>>>>>> Hi,
>>>>>> The numbers correspond to the dimensions of the ROI rectangle I drag.
>>>>>> I am pretty sure it's that, as the exception modifies with my changes...
>>>>>> and goes away when I create a perfect square.
>>>>>> For example:
>>>>>> Exception: X and Y arrays must be the same shape--got (38,)
>>>>>> and (22,).
>>>>>>
>>>>>>
>>>>>> On Tuesday, May 19, 2020 at 5:26:24 PM UTC+2, megan wrote:
>>>>>>>
>>>>>>> Hi,
>>>>>>>
>>>>>>> That is an error that occurs when you attempt to plot data that has
>>>>>>> a different number of x and y values. I need more information in order
>>>>>>> to
>>>>>>> understand why you're getting that error though. What are the shapes in
>>>>>>> the
>>>>>>> error? Can you post the whole traceback?
>>>>>>>
>>>>>>> Megan
>>>>>>>
>>>>>>> On Tue, May 19, 2020 at 9:17 AM Lavinia B <[email protected]> wrote:
>>>>>>>
>>>>>>>> Hi,
>>>>>>>> I have an ImageView in which I set data with mode=rgba. When I try
>>>>>>>> to use the ROI , the plot gives me warnings if the ROI is not
>>>>>>>> perfectly
>>>>>>>> square.
>>>>>>>> "..\src\pyqtgraph\pyqtgraph\graphicsItems\PlotCurveItem.py", line
>>>>>>>> 385, in updateData
>>>>>>>> raise Exception("X and Y arrays must be the same shape--got %s
>>>>>>>> and %s." % (self.xData.shape, self.yData.shape))
>>>>>>>>
>>>>>>>> I don't understand why is this enforced. Can somebody explain?
>>>>>>>>
>>>>>>>> Note: this does not happen for mono images, or if I keep my
>>>>>>>> selection shape square.
>>>>>>>>
>>>>>>>>
>>>>>>>> --
>>>>>>>> You received this message because you are subscribed to the Google
>>>>>>>> Groups "pyqtgraph" group.
>>>>>>>> To unsubscribe from this group and stop receiving emails from it,
>>>>>>>> send an email to [email protected].
>>>>>>>> To view this discussion on the web visit
>>>>>>>> https://groups.google.com/d/msgid/pyqtgraph/03453f45-80b6-4e4c-8589-0ca26e131205%40googlegroups.com
>>>>>>>>
>>>>>>>> <https://groups.google.com/d/msgid/pyqtgraph/03453f45-80b6-4e4c-8589-0ca26e131205%40googlegroups.com?utm_medium=email&utm_source=footer>
>>>>>>>> .
>>>>>>>>
>>>>>>> --
>>>>>> You received this message because you are subscribed to the Google
>>>>>> Groups "pyqtgraph" group.
>>>>>> To unsubscribe from this group and stop receiving emails from it,
>>>>>> send an email to [email protected].
>>>>>> To view this discussion on the web visit
>>>>>> https://groups.google.com/d/msgid/pyqtgraph/5c8c6bbd-3989-42d6-91e6-fa7dc729d527%40googlegroups.com
>>>>>>
>>>>>> <https://groups.google.com/d/msgid/pyqtgraph/5c8c6bbd-3989-42d6-91e6-fa7dc729d527%40googlegroups.com?utm_medium=email&utm_source=footer>
>>>>>> .
>>>>>>
>>>>> --
>>>> You received this message because you are subscribed to the Google
>>>> Groups "pyqtgraph" group.
>>>> To unsubscribe from this group and stop receiving emails from it, send
>>>> an email to [email protected].
>>>> To view this discussion on the web visit
>>>> https://groups.google.com/d/msgid/pyqtgraph/22b1cce2-a9da-45e2-849d-bd4271a240a3%40googlegroups.com
>>>>
>>>> <https://groups.google.com/d/msgid/pyqtgraph/22b1cce2-a9da-45e2-849d-bd4271a240a3%40googlegroups.com?utm_medium=email&utm_source=footer>
>>>> .
>>>>
>>>
--
You received this message because you are subscribed to the Google Groups
"pyqtgraph" group.
To unsubscribe from this group and stop receiving emails from it, send an email
to [email protected].
To view this discussion on the web visit
https://groups.google.com/d/msgid/pyqtgraph/265e1ccc-a989-4420-b338-1c2a1b423655o%40googlegroups.com.
