x.ptp() -> np.ptp(x)

src/kima/pykima/analysis.py

@@ -302,7 +302,7 @@ def aliases(P):
     return alias_year, alias_solar_day, alias_sidereal_day
 
 def FIP(results, oversampling=5, plot=True, adjust_oversampling=True):
-    Tobs = results.t.ptp()
+    Tobs = np.ptp(results.t)
     Dw = 2 * np.pi / Tobs
     a, b = results.priors['Pprior'].support()
 
@@ -341,7 +341,7 @@ def FIP_count_detections(results, alpha=0.05, Ptrue=None):
         from interval import Interval
         bins, fip = FIP(results, plot=False)
         ftrue = 1 / Ptrue
-        Tobs = results.t.ptp()
+        Tobs = np.ptp(results.t)
         ffip = 1 / bins[fip.argmin()]
         return ftrue in Interval(ffip - 1 / Tobs, ffip + 1 / Tobs)
 
@@ -866,9 +866,9 @@ def detection_limits(results, star_mass: Union[float, Tuple] = 1.0,
             kwargs = dict(ls='--', lw=2, alpha=0.5, zorder=-1, color='C5')
             if isinstance(res, list):
                 for r in res:
-                    ax.axvline(r.data.t.ptp(), 0.5, 1, **kwargs)
+                    ax.axvline(np.ptp(r.data.t), 0.5, 1, **kwargs)
             else:
-                ax.axvline(res.data.t.ptp(), 0.5, 1, **kwargs)
+                ax.axvline(np.ptp(res.data.t), 0.5, 1, **kwargs)
 
         # ax.hlines(s.max, bins_start, bins_end, lw=2)
         # ax.loglog(s99.bins, s99.max * mjup2mearth)

src/kima/pykima/display.py

@@ -323,7 +323,7 @@ def plot_posterior_period(res,
         ax.axvline(x=year, color='r', label='1 year', **kwline)
 
     if show_timespan:  # mark the timespan of the data
-        ax.axvline(x=res.data.t.ptp(), color='k', label='time span', **kwline)
+        ax.axvline(x=np.ptp(res.data.t), color='k', label='time span', **kwline)
 
     if show_aliases is not None:  # mark daily and yearly aliases of top peak
         from .analysis import aliases
@@ -452,7 +452,7 @@ def plot_PKE(res, mask=None, include_known_object=False, show_prior=False,
             K = K.ravel()
             E = E.ravel()
 
-        if K.size > 1 and K.ptp() > Khdr_threshold:
+        if K.size > 1 and np.ptp(K) > Khdr_threshold:
             ax1.loglog(P, K, '.', **kw)
         else:
             ax1.semilogx(P, K, '.', **kw)
@@ -461,7 +461,7 @@ def plot_PKE(res, mask=None, include_known_object=False, show_prior=False,
 
 
     else:
-        if K.size > 1 and K.ptp() > 30:
+        if K.size > 1 and np.ptp(K) > 30:
             cm = ax1.hexbin(P[P != 0.0], K[P != 0.0], gridsize=gridsize, bins='log', xscale='log',
                             yscale='log', cmap=plt.get_cmap('coolwarm_r'))
         else:
@@ -1198,7 +1198,7 @@ def corner_planet_parameters(res, fig=None, Np=None, true_values=None, period_ra
 
     # # set the parameter ranges to include everything
     # def r(x, over=0.2):
-    #     return x.min() - over * x.ptp(), x.max() + over * x.ptp()
+    #     return x.min() - over * np.ptp(x), x.max() + over * np.ptp(x)
 
 
 
@@ -2490,7 +2490,7 @@ def phase_plot(res,
         axp = fig.add_subplot(gs[:, -1])
         from astropy.timeseries import LombScargle
         gls = LombScargle(res.data.t, residuals, res.data.e)
-        freq, power = gls.autopower(maximum_frequency=1.0, minimum_frequency=1/res.data.t.ptp())
+        freq, power = gls.autopower(maximum_frequency=1.0, minimum_frequency=1/np.ptp(res.data.t))
         axp.semilogy(power, 1 / freq, 'k', alpha=0.6)
 
         kwl = dict(color='k', alpha=0.2, ls='--')
@@ -2798,7 +2798,7 @@ def plot_random_samples_multiseries(res, ncurves=50, samples=None, over=0.1, ntt
         t -= 24e5
         M0_epoch -= 24e5
 
-    tt = np.linspace(t.min() - over * t.ptp(), t.max() + over * t.ptp(), ntt + int(100 * over))
+    tt = np.linspace(t.min() - over * np.ptp(t), t.max() + over * np.ptp(t), ntt + int(100 * over))
 
     if res.has_gp:
         # let's be more reasonable for the number of GP prediction points
@@ -2808,7 +2808,7 @@ def plot_random_samples_multiseries(res, ncurves=50, samples=None, over=0.1, ntt
         kde = gaussian_kde(t)
         ttGP = kde.resample(25000 + t.size * 3).reshape(-1)
         # constrain ttGP within observed times, to not waste
-        ttGP = (ttGP + t[0]) % t.ptp() + t[0]
+        ttGP = (ttGP + t[0]) % np.ptp(t) + t[0]
         ttGP = np.r_[ttGP, t]
         ttGP.sort()  # in-place
 
@@ -2933,7 +2933,7 @@ def plot_random_samples_multiseries(res, ncurves=50, samples=None, over=0.1, ntt
             for iko in range(res.nKO):
                 KOpl = res.eval_model(samples[i], tt,
                                       single_planet=-iko - 1)[0]
-                ax1.plot(tt, KOpl - y_offset + (iko + 1) * res.data.y.ptp(),
+                ax1.plot(tt, KOpl - y_offset + (iko + 1) * np.ptp(res.data.y),
                          color='g', alpha=alpha)
 
         if res.has_gp:

src/kima/pykima/results.py

@@ -104,7 +104,7 @@ def __repr__(self):
 
     @property
     def tmiddle(self):
-        return self.t.min() + 0.5 * self.t.ptp()
+        return self.t.min() + 0.5 * np.ptp(self.t)
 
 
 @dataclass
@@ -1835,16 +1835,16 @@ def _get_tt(self, N=1000, over=0.1):
         Create array for model prediction plots. This simply returns N
         linearly-spaced times from t[0]-over*Tspan to t[-1]+over*Tspan.
         """
-        start = self.data.t.min() - over * self.data.t.ptp()
-        end = self.data.t.max() + over * self.data.t.ptp()
+        start = self.data.t.min() - over * np.ptp(self.data.t)
+        end = self.data.t.max() + over * np.ptp(self.data.t)
         return np.linspace(start, end, N)
 
     def _get_ttGP(self, N=1000, over=0.1):
         """ Create array of times for GP prediction plots. """
         kde = gaussian_kde(self.data.t)
         ttGP = kde.resample(N - self.data.N).reshape(-1)
         # constrain ttGP within observed times, to not waste
-        ttGP = (ttGP + self.data.t[0]) % self.data.t.ptp() + self.data.t[0]
+        ttGP = (ttGP + self.data.t[0]) % np.ptp(self.data.t) + self.data.t[0]
         # add the observed times as well
         ttGP = np.r_[ttGP, self.data.t]
         ttGP.sort()  # in-place
@@ -2623,7 +2623,7 @@ def _offset_times(self):
     def data_properties(self):
         t = self.data.t
         prop = {
-            'time span': (t.ptp(), 'days', True),
+            'time span': (np.ptp(t), 'days', True),
             'mean time gap': (np.ediff1d(t).mean(), 'days', True),
             'median time gap': (np.median(np.ediff1d(t)), 'days', True),
             'shortest time gap': (np.ediff1d(t).min(), 'days', True),

src/kima/pykima/utils.py

@@ -597,16 +597,16 @@ def priors_latex(results, title='Priors', label='tab:1'):
         2 * np.pi: r'$2\pi$',
         results.y.min(): r'$\min v$',
         results.y.max(): r'$\max v$',
-        results.t.ptp(): r'$\Delta t$',
-        results.y.ptp(): r'$\Delta RV$',
-        -results.y.ptp(): r'$-\Delta RV$'
+        np.ptp(results.t): r'$\Delta t$',
+        np.ptp(results.y): r'$\Delta RV$',
+        -np.ptp(results.y): r'$-\Delta RV$'
     }
     if results.model == 'RVFWHMmodel':
         translation.update({
             results.y2.min(): r'$\min$ FWHM',
             results.y2.max(): r'$\max$ FWHM',
-            results.y2.ptp(): r'$\Delta$ FWHM',
-            -results.y2.ptp(): r'$-\Delta$ FWHM'
+            np.ptp(results.y2): r'$\Delta$ FWHM',
+            -np.ptp(results.y2): r'$-\Delta$ FWHM'
         })
 
     def translate_value(value):