Merge pull request #639 from larrybradley/np114_style

Fix doctests for numpy 1.14 style changes
astropy · Feb 1, 2018 · 5a06059 · 5a06059
2 parents 572ef6f + 3806ff6
commit 5a06059
Show file tree

Hide file tree

Showing 11 changed files with 304 additions and 212 deletions.
diff --git a/docs/aperture.rst b/docs/aperture.rst
@@ -122,12 +122,13 @@ with the data and the apertures::
     >>> from photutils import aperture_photometry
     >>> data = np.ones((100, 100))
     >>> phot_table = aperture_photometry(data, apertures)
-    >>> print(phot_table)    # doctest: +SKIP
-     id xcenter ycenter  aperture_sum
+    >>> phot_table['aperture_sum'].info.format = '%.8g'  # for consistent table output
+    >>> print(phot_table)
+     id xcenter ycenter aperture_sum
           pix     pix
-    --- ------- ------- -------------
-      1    30.0    30.0 28.2743338823
-      2    40.0    40.0 28.2743338823
+    --- ------- ------- ------------
+      1    30.0    30.0    28.274334
+      2    40.0    40.0    28.274334
 
 This function returns the results of the photometry in an Astropy
 `~astropy.table.QTable`.  In this example, the table has four columns,
@@ -159,7 +160,7 @@ by a factor of 5 (``subpixels=5``) in each dimension::
 
     >>> phot_table = aperture_photometry(data, apertures, method='subpixel',
     ...                                  subpixels=5)
-    >>> print(phot_table)    # doctest: +SKIP
+    >>> print(phot_table)
      id xcenter ycenter aperture_sum
           pix     pix
     --- ------- ------- ------------
@@ -192,12 +193,14 @@ Suppose that we wish to use three circular apertures, with radii of 3,
     >>> radii = [3., 4., 5.]
     >>> apertures = [CircularAperture(positions, r=r) for r in radii]
     >>> phot_table = aperture_photometry(data, apertures)
-    >>> print(phot_table)    # doctest: +SKIP
+    >>> for col in phot_table.colnames:
+    ...     phot_table[col].info.format = '%.8g'  # for consistent table output
+    >>> print(phot_table)
      id xcenter ycenter aperture_sum_0 aperture_sum_1 aperture_sum_2
           pix     pix
     --- ------- ------- -------------- -------------- --------------
-      1    30.0    30.0  28.2743338823  50.2654824574  78.5398163397
-      2    40.0    40.0  28.2743338823  50.2654824574  78.5398163397
+      1      30      30      28.274334      50.265482      78.539816
+      2      40      40      28.274334      50.265482      78.539816
 
 For multiple apertures, the output table column names are appended
 with the ``positions`` index.
@@ -212,12 +215,14 @@ specify ``a``, ``b``, and ``theta``::
     >>> theta = np.pi / 4.
     >>> apertures = EllipticalAperture(positions, a, b, theta)
     >>> phot_table = aperture_photometry(data, apertures)
-    >>> print(phot_table)    # doctest: +SKIP
-     id xcenter ycenter  aperture_sum
+    >>> for col in phot_table.colnames:
+    ...     phot_table[col].info.format = '%.8g'  # for consistent table output
+    >>> print(phot_table)
+     id xcenter ycenter aperture_sum
           pix     pix
-    --- ------- ------- -------------
-      1    30.0    30.0 47.1238898038
-      2    40.0    40.0 47.1238898038
+    --- ------- ------- ------------
+      1      30      30     47.12389
+      2      40      40     47.12389
 
 Again, for multiple apertures one should input a list of aperture
 objects, each with identical positions::
@@ -228,12 +233,14 @@ objects, each with identical positions::
     >>> apertures = [EllipticalAperture(positions, a=ai, b=bi, theta=theta)
     ...              for (ai, bi) in zip(a, b)]
     >>> phot_table = aperture_photometry(data, apertures)
-    >>> print(phot_table)    # doctest: +SKIP
+    >>> for col in phot_table.colnames:
+    ...     phot_table[col].info.format = '%.8g'  # for consistent table output
+    >>> print(phot_table)
      id xcenter ycenter aperture_sum_0 aperture_sum_1 aperture_sum_2
           pix     pix
     --- ------- ------- -------------- -------------- --------------
-      1    30.0    30.0  47.1238898038  75.3982236862  109.955742876
-      2    40.0    40.0  47.1238898038  75.3982236862  109.955742876
+      1      30      30       47.12389      75.398224      109.95574
+      2      40      40       47.12389      75.398224      109.95574
 
 
 Background Subtraction
@@ -267,12 +274,14 @@ We then perform the photometry in both apertures::
 
     >>> apers = [apertures, annulus_apertures]
     >>> phot_table = aperture_photometry(data, apers)
-    >>> print(phot_table)    # doctest: +SKIP
+    >>> for col in phot_table.colnames:
+    ...     phot_table[col].info.format = '%.8g'  # for consistent table output
+    >>> print(phot_table)
      id xcenter ycenter aperture_sum_0 aperture_sum_1
           pix     pix
     --- ------- ------- -------------- --------------
-      1    30.0    30.0  28.2743338823  87.9645943005
-      2    40.0    40.0  28.2743338823  87.9645943005
+      1      30      30      28.274334      87.964594
+      2      40      40      28.274334      87.964594
 
 Note that we cannot simply subtract the aperture sums because the
 apertures have different areas.
@@ -289,11 +298,12 @@ background times the circular aperture area::
     >>> bkg_sum = bkg_mean * apertures.area()
     >>> final_sum = phot_table['aperture_sum_0'] - bkg_sum
     >>> phot_table['residual_aperture_sum'] = final_sum
-    >>> print(phot_table['residual_aperture_sum'])    # doctest: +FLOAT_CMP
+    >>> phot_table['residual_aperture_sum'].info.format = '%.8g'  # for consistent table output
+    >>> print(phot_table['residual_aperture_sum'])
     residual_aperture_sum
     ---------------------
-        -7.1054273576e-15
-        -7.1054273576e-15
+           -7.1054274e-15
+           -7.1054274e-15
 
 The result here should be zero because all of the data values are 1.0
 (the tiny difference from 0.0 is due to numerical precision).
@@ -316,12 +326,14 @@ pixel's value and saved it in the array ``error``::
 
     >>> error = 0.1 * data
     >>> phot_table = aperture_photometry(data, apertures, error=error)
-    >>> print(phot_table)    # doctest: +SKIP
-     id xcenter ycenter  aperture_sum aperture_sum_err
+    >>> for col in phot_table.colnames:
+    ...     phot_table[col].info.format = '%.8g'  # for consistent table output
+    >>> print(phot_table)
+     id xcenter ycenter aperture_sum aperture_sum_err
           pix     pix
-    --- ------- ------- ------------- ----------------
-      1    30.0    30.0 28.2743338823   0.531736155272
-      2    40.0    40.0 28.2743338823   0.531736155272
+    --- ------- ------- ------------ ----------------
+      1      30      30    28.274334       0.53173616
+      2      40      40    28.274334       0.53173616
 
 ``'aperture_sum_err'`` values are given by:
 
@@ -363,18 +375,20 @@ photometry by providing an image mask via the ``mask`` keyword::
     >>> data[2, 2] = 100.   # bad pixel
     >>> mask[2, 2] = True
     >>> t1 = aperture_photometry(data, aperture, mask=mask)
-    >>> print(t1['aperture_sum'])    # doctest: +FLOAT_CMP
-     aperture_sum
-     -------------
-     11.5663706144
+    >>> t1['aperture_sum'].info.format = '%.8g'  # for consistent table output
+    >>> print(t1['aperture_sum'])
+    aperture_sum
+    ------------
+       11.566371
 
 The result is very different if a ``mask`` image is not provided::
 
     >>> t2 = aperture_photometry(data, aperture)
-    >>> print(t2['aperture_sum'])    # doctest: +FLOAT_CMP
+    >>> t2['aperture_sum'].info.format = '%.8g'  # for consistent table output
+    >>> print(t2['aperture_sum'])
     aperture_sum
-    -------------
-    111.566370614
+    ------------
+       111.56637
 
 
 Aperture Photometry Using Sky Coordinates

diff --git a/docs/background.rst b/docs/background.rst
@@ -76,18 +76,18 @@ background level of 5::
 
     >>> import numpy as np
     >>> from astropy.stats import biweight_location
-    >>> print(np.median(data))
-    5.2255295184
-    >>> print(biweight_location(data))
-    5.1867597555
+    >>> print(np.median(data))  # doctest: +FLOAT_CMP
+    5.225529518399048
+    >>> print(biweight_location(data))  # doctest: +FLOAT_CMP
+    5.186759755495727
 
 Similarly, using the median absolute deviation to estimate the
 background noise level gives a value that is larger than the true
 value of 2::
 
     >>> from astropy.stats import mad_std
-    >>> print(mad_std(data))    # doctest: +FLOAT_CMP
-    2.1443728009
+    >>> print(mad_std(data))  # doctest: +FLOAT_CMP
+    2.1443760096598914
 
 
 Sigma Clipping Sources
@@ -103,8 +103,8 @@ levels::
 
     >>> from astropy.stats import sigma_clipped_stats
     >>> mean, median, std = sigma_clipped_stats(data, sigma=3.0, iters=5)
-    >>> print((mean, median, std))    # doctest: +FLOAT_CMP
-    (5.1991386516217908, 5.1555874333582912, 2.0942752121329691)
+    >>> print((mean, median, std))  # doctest: +FLOAT_CMP
+    (5.199138651621793, 5.155587433358291, 2.094275212132969)
 
 
 Masking Sources
@@ -132,8 +132,8 @@ source detections and dilate using a 11x11 box:
     >>> from photutils import make_source_mask
     >>> mask = make_source_mask(data, snr=2, npixels=5, dilate_size=11)
     >>> mean, median, std = sigma_clipped_stats(data, sigma=3.0, mask=mask)
-    >>> print((mean, median, std))    # doctest: +FLOAT_CMP
-    (5.0010134754755695, 5.0005849056043763, 1.970887100626572)
+    >>> print((mean, median, std))  # doctest: +FLOAT_CMP
+    (5.001013475475569, 5.000584905604376, 1.970887100626572)
 
 Of course, the source detection and masking procedure can be iterated
 further.  Even with one iteration we are within 0.02% of the true
@@ -229,7 +229,7 @@ background gradient to the image defined above::
     >>> y, x = np.mgrid[:ny, :nx]
     >>> gradient =  x * y / 5000.
     >>> data2 = data + gradient
-    >>> plt.imshow(data2, norm=norm, origin='lower', cmap='Greys_r')    # doctest: +SKIP
+    >>> plt.imshow(data2, norm=norm, origin='lower', cmap='Greys_r')  # doctest: +SKIP
 
 .. plot::
 
@@ -271,10 +271,10 @@ attributes, respectively:
 
 .. doctest-requires:: scipy
 
-    >>> print(bkg.background_median)
-    10.8219978626
-    >>> print(bkg.background_rms_median)
-    2.29882053968
+    >>> print(bkg.background_median)  # doctest: +FLOAT_CMP
+    10.821997862561792
+    >>> print(bkg.background_rms_median)  # doctest: +FLOAT_CMP
+    2.298820539683762
 
 Let's plot the background image:
 
@@ -348,8 +348,8 @@ Let's create such an image and plot it (NOTE: this example requires
 
     >>> from scipy.ndimage import rotate
     >>> data3 = rotate(data2, -45.)
-    >>> norm = ImageNormalize(stretch=SqrtStretch())    # doctest: +SKIP
-    >>> plt.imshow(data3, origin='lower', cmap='Greys_r', norm=norm)    # doctest: +SKIP
+    >>> norm = ImageNormalize(stretch=SqrtStretch())  # doctest: +SKIP
+    >>> plt.imshow(data3, origin='lower', cmap='Greys_r', norm=norm)  # doctest: +SKIP
 
 .. plot::
 
@@ -386,8 +386,8 @@ apply the coverage mask to the returned background image:
 .. doctest-requires:: scipy
 
     >>> back3 = bkg3.background * ~mask
-    >>> norm = ImageNormalize(stretch=SqrtStretch())    # doctest: +SKIP
-    >>> plt.imshow(back3, origin='lower', cmap='Greys_r', norm=norm)    # doctest: +SKIP
+    >>> norm = ImageNormalize(stretch=SqrtStretch())  # doctest: +SKIP
+    >>> plt.imshow(back3, origin='lower', cmap='Greys_r', norm=norm)  # doctest: +SKIP
 
 .. plot::
 

diff --git a/docs/detection.rst b/docs/detection.rst
@@ -62,20 +62,23 @@ above the background. Running this class on the data yields an astropy
     >>> from photutils import DAOStarFinder
     >>> daofind = DAOStarFinder(fwhm=3.0, threshold=5.*std)    # doctest: +REMOTE_DATA
     >>> sources = daofind(data - median)    # doctest: +REMOTE_DATA
+    >>> for col in sources.colnames:    # doctest: +REMOTE_DATA
+    ...     sources[col].info.format = '%.8g'  # for consistent table output
     >>> print(sources)    # doctest: +REMOTE_DATA
-     id   xcentroid     ycentroid   ...  peak       flux           mag
-    --- ------------- ------------- ... ------ ------------- ---------------
-      1 144.247567164 6.37979042704 ... 6903.0 5.70143033038  -1.88995955438
-      2 208.669068628 6.82058053777 ... 7896.0 6.72306730455  -2.06891864748
-      3 216.926136655  6.5775933198 ... 2195.0 1.66737467591 -0.555083002864
-      4 351.625190383  8.5459013233 ... 6977.0 5.90092548147  -1.92730032571
-      5 377.519909958 12.0655009987 ... 1260.0 1.11856203781 -0.121650189969
-    ...           ...           ... ...    ...           ...             ...
-    281 268.049236979 397.925371446 ... 9299.0 6.22022587541  -1.98451538884
-    282 268.475068392 398.020998272 ... 8754.0 6.05079160593  -1.95453048936
-    283  299.80943822 398.027911813 ... 8890.0 6.11853416663  -1.96661847383
-    284 315.689448343  398.70251891 ... 6485.0 5.55471107793  -1.86165368631
-    285 360.437243037 398.698539555 ... 8079.0 5.26549321379  -1.80359764345
+     id xcentroid ycentroid sharpness  ... sky peak    flux       mag
+    --- --------- --------- ---------- ... --- ---- --------- ------------
+      1 144.24757 6.3797904 0.58156257 ...   0 6903 5.7014303   -1.8899596
+      2 208.66907 6.8205805 0.48348966 ...   0 7896 6.7230673   -2.0689186
+      3 216.92614 6.5775933 0.69359525 ...   0 2195 1.6673747    -0.555083
+      4 351.62519 8.5459013 0.48577834 ...   0 6977 5.9009255   -1.9273003
+      5 377.51991 12.065501 0.52038488 ...   0 1260  1.118562  -0.12165019
+    ...       ...       ...        ... ... ...  ...       ...          ...
+    280 345.59306 395.38222   0.384078 ...   0 9350 5.0559084    -1.759498
+    281 268.04924 397.92537 0.29650715 ...   0 9299 6.2202259   -1.9845154
+    282 268.47507   398.021 0.28325741 ...   0 8754 6.0507916   -1.9545305
+    283 299.80944 398.02791 0.32011339 ...   0 8890 6.1185342   -1.9666185
+    284 315.68945 398.70252 0.29502138 ...   0 6485 5.5547111   -1.8616537
+    285 360.43724 398.69854 0.81147144 ...   0 8079 5.2654932   -1.8035976
     Length = 285 rows
 
 Let's plot the image and mark the location of detected sources:
@@ -143,19 +146,20 @@ sigma above the background and a separated by at least 2 pixels:
     >>> mean, median, std = sigma_clipped_stats(data, sigma=3.0)
     >>> threshold = median + (10.0 * std)
     >>> tbl = find_peaks(data, threshold, box_size=5)
+    >>> tbl['peak_value'].info.format = '%.8g'  # for consistent table output
     >>> print(tbl[:10])    # print only the first 10 peaks
-    x_peak y_peak   peak_value
-    ------ ------ -------------
-       233      0 27.4778521972
-       236      1  27.339519624
-       289     22 35.8532759965
-       442     31 30.2399941373
-         1     40 35.5482863002
-        89     59 41.2190469279
-         7     70 33.2880647048
-       258     75 26.5624808518
-       463     80 28.7588206692
-       182     93 38.0885687202
+    x_peak y_peak peak_value
+    ------ ------ ----------
+       233      0  27.477852
+       236      1   27.33952
+       289     22  35.853276
+       442     31  30.239994
+         1     40  35.548286
+        89     59  41.219047
+         7     70  33.288065
+       258     75  26.562481
+       463     80  28.758821
+       182     93  38.088569
 
 And let's plot the location of the detected peaks in the image: