DRS-Sum_to_oxyde and MAP

ddsavard

Hi guys,
I'm working on quantitative method using the DRS for the summation to oxydes and normalized. We would like to improve our mapping quantification using that technique, so it would be possible to produce suitable standard-less quantitative map of multiple mineral phases on a single image, for example Apatite, Sulfide, Magnetite and Pyroxene all together. When using the inspector tool to select individual mineral area, the results are very good, the balance of elements (re-converted to oxyde in EXCEL), is really close to 100%. However, the scale from the images are completely wrong by an order of magnitude, sometime showing concentrations of more than 100%... how is concentrations on the images differ from inspector tool, I mean, each pixel is recalculated to 100%, right?
Thank you
Dany

ddsavard

..is it possible that negative value creates the issue? Is it possible to replace negative values by 0 directly in the DRS?

Joe

Hi Dany,

Can you share an example with the support email? I'll have a look.

You could replace negative values from python with something like:

for name in data.timeSeriesNames(data.Output):
    channel = data.timeSeries(name)
    channel.data()[channel.data() < 0] = 0

All the best,

ddsavard

Thanks Joe, I already had this tool to remove the negative for data input. I will try that one with output. Thanks

ddsavard

I got it Joe!
First the zero remover from the output seem to work as well as the one for the input. Thanks for that. But that didn't solved the problem. The problem was because I used a custom filters on the images for those experiment. Using the "median" filter, the data looks ok, while using the custom is clearly the source of the increase I observed. So I assumed that filters are applied after calculation, so the scale is wrongly re-adjusted.
I will keep an eye on the median filter, otherwise I will not use filter at all.
Thank you!
Dany