PR for #22: deploy shiny (#23)

* new shiny page for #22

* finishes for #22

* changes behavior for `.mat` files for #22

* Adds pointer to readme for #22

* Update shiny/app.py

Co-authored-by: jmshapir <[email protected]>

* Update shiny/app.py

Co-authored-by: jmshapir <[email protected]>

* changes for interface for #22

* Update shiny/app.py

Co-authored-by: jmshapir <[email protected]>

* reverts version for Bootstrapreport for #22

* boosts compatibility with matlab for #22

* extends compatibility to RDATA for #22

* fix for #22

* Fix for #23

* overhaul for #22

* Wording change for #23

@MosesStewart fyi

* added stable link to readme for #22

* Tidying for #22

* Version for #23

---------

Co-authored-by: jmshapir <[email protected]>

Author: MosesStewart

README.md

@@ -8,5 +8,8 @@ Documentation:
 * [Methods](./BootstrapReport.lyx)
 * [Examples](./examples/)
 
+WebApp:
+* [ShinyBootstrapReport](https://jmslab.shinyapps.io/BootstrapReport)
+
 Citation:
 * Andrews, Isaiah and Jesse M. Shapiro. "Bootstrap Diagnostics for Irregular Estimators." 2023. Working Paper.

examples/_generate_example_data.py

@@ -37,18 +37,17 @@ def generate_example_dataset(name):
         bootstrap replicates, or 'gamma' for gamma mean bootstrap replicates
     :return: Replicates and estimate for the data
     """
-    np.random.seed(10042002)
     rng = np.random.default_rng(seed = 10042002)
     if name == 'normal':
         num_rep, mean, std = 100, 0.74, 0.086
-        raw_data = norm.rvs(size = num_rep, loc = mean, scale = np.sqrt(num_rep) * std)
+        raw_data = rng.normal(size = num_rep, loc = mean, scale = np.sqrt(num_rep) * std)
         replicates = get_replicates(raw_data)
         estimate = np.median(raw_data)
     elif name == 'gamma':
         num_rep, variance, mean = 499, 0.186, 2.189
-        raw_data = gamma.rvs((mean**2)/(variance * num_rep), scale = (variance * num_rep)/mean, size = 50)
+        raw_data = rng.gamma(shape = (mean**2)/(variance * num_rep), scale = (variance * num_rep)/mean, size = 499)
         replicates = get_replicates(raw_data, num_rep = num_rep)
-        estimate = np.average(raw_data, weights = uniform.rvs(size = 50))
+        estimate = np.average(raw_data, weights = rng.uniform(size = 499))
     elif name == 'ratio':
         num_rep, numer_std, mean = 1000, 1, 0
         replicates = rng.normal(size = num_rep, loc = mean, scale = numer_std)/ \

examples/gamma_estimate.csv

@@ -1,2 +1,2 @@
 estimate,std_err
-3.329153904463066,2.0256286444979983
+1.8450286322049927,0.3898786195682134