Added exercise solutions and added additional reading references to README

Author: Christopher Fonnesbeck

README.md

@@ -77,6 +77,20 @@ Otherwise, another easy way to install all the necessary packages is to use Cont
 
 For those requiring extra assistance with installing packages, there will be a help desk available Monday, June 24, 7-8AM at the conference venue.
 
+## Additional Reading
+
+[The Ecological Detective: Confronting Models with Data](http://www.amazon.com/Ecological-Detective-Confronting-Models-Data/dp/0691034974/ref=sr_1_1?s=books&ie=UTF8&qid=1372250186&sr=1-1&keywords=ecological+detective), Ray Hilborn and Marc Mangel
+
+[Data Analysis Using Regression and Multilevel/Hierarchical Models](http://www.amazon.com/Analysis-Regression-Multilevel-Hierarchical-Models/dp/052168689X/ref=sr_1_1?s=books&ie=UTF8&qid=1372250274&sr=1-1&keywords=gelman), Andrew Gelman and Jennifer Hill
+
+[The Elements of Statistical Learning](http://www-stat.stanford.edu/~tibs/ElemStatLearn/), Hastie, Tibshirani and Friedman
+
+[A First Course in Bayesian Statistical Methods](http://www.amazon.com/Bayesian-Statistical-Methods-Springer-Statistics/dp/1441928286/ref=sr_1_24?ie=UTF8&qid=1372250835&sr=8-24&keywords=bayesian), Peter Hoff
+
+[Regression Modeling Strategies](http://www.amazon.com/Regression-Modeling-Strategies-Applications-Statistics/dp/1441929185/ref=sr_1_1?s=books&ie=UTF8&qid=1372250898&sr=1-1&keywords=harrell+regression), Frank Harrell
+
+[Python for Data Analysis](http://shop.oreilly.com/product/0636920023784.do), Wes McKinney
+
 [1]: http://nbviewer.ipython.org/urls/gist.github.com/fonnesbeck/5850375/raw/c18cfcd9580d382cb6d14e4708aab33a0916ff3e/1.+Introduction+to+Pandas.ipynb   "Introduction to Pandas"
 [2]: http://nbviewer.ipython.org/urls/gist.github.com/fonnesbeck/5850413/raw/3a9406c73365480bc58d5e75bc80f7962243ba17/2.+Data+Wrangling+with+Pandas.ipynb "Data wrangling with Pandas"
 [3]: http://nbviewer.ipython.org/urls/gist.github.com/fonnesbeck/5850463/raw/a29d9ffb863bfab09ff6c1fc853e1d5bf69fe3e4/3.+Plotting+and+Visualization.ipynb "Plotting and visualization"

exercise_solutions.md

@@ -0,0 +1,65 @@
+## Exercise Solutions
+
+General polynomial function:
+
+    def calc_poly(params, data):
+        x = np.c_[[data**i for i in range(len(params))]]
+        return np.dot(params, x)
+
+Microbiome exercise:
+
+    metadata = pd.read_excel('data/microbiome/metadata.xls', sheetname='Sheet1')
+
+    chunks = []
+    for i in range(9):
+        this_file = pd.read_excel('data/microbiome/MID{0}.xls'.format(i+1), 'Sheet 1', index_col=0, header=None, names=['Taxon', 'Count'])
+        this_file.columns = ['Count']
+        this_file.index.name = 'Taxon'
+        for m in metadata.columns:
+            this_file[m] = metadata.ix[i][m]
+        chunks.append(this_file)
+
+    pd.concat(chunks)
+
+Titanic proportions:
+
+    titanic = pd.read_excel("data/titanic.xls", "titanic")
+
+    titanic.groupby('sex')['survived'].mean()
+
+    titanic.groupby(['pclass','sex'])['survived'].mean()
+
+    titanic['agecat'] = pd.cut(titanic.age, [0, 13, 20, 64, 100], labels=['child', 'adolescent', 'adult', 'senior'])
+    titanic.groupby(['agecat', 'pclass','sex'])['survived'].mean()
+
+Survivor KDE plots:
+
+    surv = dict(list(titanic.groupby('survived')))
+    for s in surv:
+        surv[s]['age'].dropna().plot(kind='kde', label=bool(s)*'survived' or 'died', grid=False)
+    legend()
+    xlim(0,100)
+
+OBP:
+
+    baseball[['h','bb', 'hbp']].sum(axis=1).div(
+            baseball[['bb', 'hbp','ab', 'sf']].sum(axis=1)
+                                ).order(ascending=False)
+
+Cervical dystonia estimation:
+
+    norm_like = lambda theta, x: -np.log(norm.pdf(x, theta[0], theta[1])).sum()
+
+    fmin(norm_like, np.array([1,2]), args=(cdystonia.twstrs[(cdystonia.obs==6) & (cdystonia.treat=='Placebo')],))
+    fmin(norm_like, np.array([1,2]), args=(cdystonia.twstrs[(cdystonia.obs==6) & (cdystonia.treat=='5000U')],))
+
+Cervical dystonia bootstrapping:
+
+    x = cdystonia.twstrs[(cdystonia.obs==6) & (cdystonia.treat=='Placebo') & (cdystonia.twstrs.notnull())].values
+    n = len(x)
+    s = [x[np.random.randint(0,n,n)].mean() for i in range(R)]
+    placebo_mean = np.sum(s)/R
+
+    s_sorted = np.sort(s)
+    alpha = 0.05
+    s_sorted[[(R+1)*alpha/2, (R+1)*(1-alpha/2)]]