fixed postprocess

augur/analyze.py

@@ -181,12 +181,14 @@ def get_derivatives(self, force=False, method='5pt_stencil'):
         else:
             return self.derivatives
 
-    def get_fisher_matrix(self, method='5pt_stencil'):
+    def get_fisher_matrix(self, method='5pt_stencil', save_txt=True):
         # Compute Fisher matrix assuming Gaussian likelihood (around self.x)
         if self.derivatives is None:
             self.get_derivatives(method=method)
         if self.Fij is None:
             self.Fij = np.einsum('il, lm, jm', self.derivatives, self.lk.inv_cov, self.derivatives)
+            if save_txt:
+                np.savetxt(self.config['output'], self.Fij)
             return self.Fij
         else:
             return self.Fij

augur/postprocess.py

@@ -3,6 +3,7 @@
 from matplotlib.collections import EllipseCollection
 import astropy.table
 from scipy.stats import norm, chi2
+from augur.utils.config_io import parse_config
 import os
 
 
@@ -16,16 +17,14 @@ def postprocess(config):
     config : dict
         The yaml parsed dictional of the input yaml file
     """
+    config = parse_config(config)
     pconfig = config["postprocess"]
-    outdir = config["analyze"]["cosmosis"]["output_dir"]
-    fid_params = config["generate"]["parameters"]
-    var_params = config["analyze"]["cosmosis"]["parameters"]
-    input = os.path.join(outdir, "chain.txt")
-    fisher = np.loadtxt(input)
+    fid_params = config["cosmo"]
+    var_params = config["fisher"]["var_pars"]
+    fisher = np.loadtxt(config["fisher"]["output"])
     npars = fisher.shape[0]
-    keys = astropy.table.Table.read(input, format="ascii").keys()
-    keys = np.array(keys)
-
+    keys = np.array(var_params)
+    outdir = config["postprocess"]["outdir"]
     if "labels" not in pconfig.keys():
         labels = keys
     else:
@@ -39,7 +38,7 @@ def postprocess(config):
         centers = np.zeros(npars)
         ind = 0
         for key in fid_params.keys():
-            if key in var_params.keys():
+            if key in var_params:
                 centers[ind] = fid_params[key]
                 ind += 1
 
@@ -90,8 +89,8 @@ def postprocess(config):
     for i in range(npairplots):
         pair0 = pairplots[2*i].split("(")[1]
         pair1 = pairplots[2*i+1].split(")")[0]
-        key1 = f"params--{pair0.lower()}"
-        key2 = f"params--{pair1.lower()}"
+        key1 = f"{pair0}"
+        key2 = f"{pair1}"
         if (key1 not in keys) or (key2 not in keys):
             # The selected set of parameters is not in the forecast
             continue
@@ -115,8 +114,8 @@ def postprocess(config):
             f.savefig(os.path.join(outdir, f"{pair0}--{pair1}.pdf"))
 
     # w0 -- wa plots are always made
-    iw = np.where(keys == "params--w0")[0][0]
-    iwa = np.where(keys == "params--wa")[0][0]
+    iw = np.where(keys == "w0")[0][0]
+    iwa = np.where(keys == "wa")[0][0]
     sig_w0 = np.sqrt(inv_cache[iw, iw])
     sig_wa = np.sqrt(inv_cache[iwa, iwa])
     FOM, FOM2 = get_FoM_all(fisher, iw, iwa, CL)

examples/config_test.yml

@@ -75,7 +75,7 @@ cov_options:
     # ------
     # Or you can also get it from a file
     cov_type : 'SRD'
-    SRD_cov_path : '../data/Y1_3x2_SRD_cov.npy'
+    SRD_cov_path : './data/Y1_3x2_SRD_cov.npy'
     # Or using TJPCov
     #cov_type : 'tjpcov'
     #IA : 0.0
@@ -98,7 +98,7 @@ fisher:
     #     n_s: [0.9, 0.9645, 1.0]
     #     #mult_bias: [-0.1 0.0 0.1]
     step: 1e-2
-    output: 'fisher.dat'
+    output: 'output/fisher.dat'
     fisher_bias:
         biased_dv: ''  # Path to file containing modified data vector with the systematic shift to probe
                     # If the file is provided it should be a FITS or ASCII file with the  same binning as the
@@ -109,4 +109,9 @@ fisher:
             lens0_delta_z: 0.01
             src3_delta_z: 0.005
 
-
+postprocess:
+  latex_table: output/latex_table.tex
+  triangle_plot: output/triangle_plot.pdf
+  pairplots: [(w0, wa), (omega_c, sigma8)]
+  outdir: 'output/'
+  CL: 0.68