Fix PSF position shift inversion, kernel accumulation, and ref baseline

psf_photom.py:
- psf_position: invert finite check so invalid (NaN/inf) ext_shift is
  zeroed rather than valid shifts being discarded
- minimize_psf_flux: convolve a local copy of self.psf with the kernel
  instead of mutating self.psf in-place; previously each optimizer
  iteration re-convolved the already-convolved PSF, progressively
  broadening it through the fit

tessreduce.py:
- reduce: replace nanmin with nanpercentile(1) for reference baseline
  removal in difference imaging mode; nanmin is dominated by a single
  outlier negative pixel (cosmic ray, bad pixel), biasing every
  difference frame by that amount

helpers.py:
- _clip_region: fix image[rx,ry] → image[ry,rx] in threshold check
  so mask is applied to the same pixels as the sigma-clipped stats
- grad_clip_fill_bkg: fix component size arrays to cover labels
  1..n_objects (exclude background label 0, include last component)

Author: CheerfulUser

tessreduce/psf_photom.py

@@ -162,7 +162,7 @@ def psf_position(self,image,error,limx=0.8,limy=0.8,ext_shift=[0,0]):#,surface=F
             error = np.ones_like(image)
         #brightloc = 
         if np.nansum(image) > 0:
-            if np.isfinite(ext_shift).all():
+            if not np.isfinite(ext_shift).all():
                 ext_shift[0] = 0; ext_shift[1] = 0
             normimage = image / np.nansum(image)    # normalise the image
 
@@ -215,10 +215,9 @@ def minimize_psf_flux(self,coeff,image,error=None,surface=True,order=2,kernel=No
             s = polynomial_surface(xx,yy,plane_coeff,order)
         else:
             s = 0
-        if kernel is not None:
-            self.psf = fftconvolve(self.psf, kernel, mode='same')
+        psf = fftconvolve(self.psf, kernel, mode='same') if kernel is not None else self.psf
 
-        res = np.nansum((image - self.psf*coeff[0] - s)**2/error)
+        res = np.nansum((image - psf*coeff[0] - s)**2/error)
         return res
 
     def psf_flux(self,image,error=None,ext_shift=None,surface=True,poly_order=3,kernel=None):

tessreduce/tessreduce.py

@@ -2763,7 +2763,7 @@ def reduce(self, aper = None, align = None, parallel = None, calibrate=None,
 					self.ref = deepcopy(self.flux[self.ref_ind])
 				elif self._ref_type.lower() == 'stack':
 					self.stack_ref()
-				self.ref -= np.nanmin(self.ref)
+				self.ref -= np.nanpercentile(self.ref, 1)
 				self.flux -= self.ref
 
 				# self.ref -= self.bkg[self.ref_ind]