def binned(self, x, y):
# pretty sure this treats nans in y correctly, but should double-check!
bin_means = binnedstat(x, y, bins=self.bin_edges,
statistic=np.nanmean)[0]
return (self.bin_edges[:-1] + self.bin_edges[1:]) / 2., bin_means
def binned(self,x,y):
# pretty sure this treats nans in y correctly, but should double-check!
bin_means = binnedstat(x,y,bins=self.bin_edges,statistic=np.nanmean)[0]
return bin_means
def binner(ls, cls, bin_edges):
x = ls.copy()
y = cls.copy()
cents = (bin_edges[:-1] + bin_edges[1:]) / 2.0
bin_edges_min = bin_edges.min()
bin_edges_max = bin_edges.max()
y[x < bin_edges_min] = 0
y[x > bin_edges_max] = 0
bin_means = binnedstat(x, y, bins=bin_edges, statistic=np.nanmean)[0]
return cents, bin_means
def bin(self, ix, iy, stat=np.nanmean):
x = ix.copy()
y = iy.copy()
# this just prevents an annoying warning (which is otherwise informative) everytime
# all the values outside the bin_edges are nans
y[x < self.bin_edges_min] = 0
y[x > self.bin_edges_max] = 0
# pretty sure this treats nans in y correctly, but should double-check!
bin_means = binnedstat(x, y, bins=self.bin_edges, statistic=stat)[0]
return self.cents, bin_means
def bin1d(bin_edges,ix,iy,stat=np.nanmean):
numbins = len(bin_edges)-1
cents = (bin_edges[:-1]+bin_edges[1:])/2.
bin_edges_min = bin_edges.min()
bin_edges_max = bin_edges.max()
x = ix.copy()
y = iy.copy()
# this just prevents an annoying warning (which is otherwise informative) everytime
# all the values outside the bin_edges are nans
y[x<bin_edges_min] = 0
y[x>bin_edges_max] = 0
# pretty sure this treats nans in y correctly, but should double-check!
bin_means = binnedstat(x,y,bins=bin_edges,statistic=stat)[0]
return cents,bin_means
def bin_annuli(ells, cls, bin_edges):
numer = binnedstat(ells, ells * cls, bins=bin_edges,
statistic=np.nanmean)[0]
denom = binnedstat(ells, ells, bins=bin_edges, statistic=np.nanmean)[0]
return numer / denom