Fixed bug that caps the limit of the indicators.

README.md

@@ -16,5 +16,5 @@ It can also be consulted in the working paper:
 * Guerrero O., Guariso D., and Castañeda G. (2023) *Aid effectiveness in sustainable development: A multidimensional approach.* World Development (https://doi.org/10.1016/j.worlddev.2023.106256).
 
 
-Previous further information on PPI and support materials visit its <a href="http://policypriority.org" target="_blank">homepage</a> .
+Previous further information on publications, impact, media coverage, and supporting materials visit its <a href="http://policypriority.org" target="_blank">homepage</a> .
 

source_code/policy_priority_inference.py

@@ -421,13 +421,13 @@ def run_ppi(I0, alphas, alphas_prime, betas, A=None, R=None, bs=None, qm=None, r
         # if theoretical maximums are provided, make sure the indicators do not surpass them
         if Imax is not None:
             with_bound = ~np.isnan(Imax)
-            newI[with_bound & (newI[with_bound] > Imax[with_bound])] = Imax[with_bound & (newI[with_bound] > Imax[with_bound])]
+            newI[with_bound & (newI > Imax)] = Imax[with_bound & (newI > Imax)]
             assert np.sum(newI[with_bound] > Imax[with_bound])==0, 'some indicators have surpassed their theoretical upper bound!'
 
         # if theoretical minimums are provided, make sure the indicators do not become lower than them
         if Imin is not None:
             with_bound = ~np.isnan(Imin)
-            newI[with_bound & (newI[with_bound] < Imin[with_bound])] = Imin[with_bound & (newI[with_bound] < Imin[with_bound])]
+            newI[with_bound & (newI < Imin)] = Imin[with_bound & (newI < Imin)]
             assert np.sum(newI[with_bound] < Imin[with_bound])==0, 'some indicators have surpassed their theoretical lower bound!'
 
         # if governance parameters are endogenous, make sure they are not larger than 1