fix white lines in Power plots

08-power.Rmd

@@ -103,8 +103,8 @@ penguins %>%
   filter(species %in% c("Adelie", "Gentoo")) %>%
   ggplot() +
   geom_histogram(aes(x = bill_length_mm, fill = species), alpha = 0.50) +
-  geom_vline(xintercept =  38.8) +
-  geom_vline(xintercept = 47.5)
+  geom_vline(xintercept =  38.8, col = "black") +
+  geom_vline(xintercept = 47.5, col = "black")
 ```
 
 In the grand scheme of things the difference between these two populations is relatively small in magnitude: their histograms overlap, indicating that some penguins in both species often have similar bill lengths. But, what if the bills of all of the Gentoo penguins *magically* grew an extra 15 mm?
@@ -116,8 +116,8 @@ penguins %>%
   mutate(bill_length_mm = if_else(species == "Gentoo", bill_length_mm + 15, bill_length_mm)) %>% 
   ggplot() + 
   geom_histogram(aes(x = bill_length_mm, fill = species), alpha = 0.56) +
-  geom_vline(xintercept =  38.8) +
-  geom_vline(xintercept = 62.5) # I found this from the new mutated values separately
+  geom_vline(xintercept =  38.8, col = "black") +
+  geom_vline(xintercept = 62.5, col = "black") # I found this from the new mutated values separately
 ```
 
 And now perform the t-test on this new *magical* data:

docs/correlation-and-simple-linear-regression.html

@@ -342,13 +342,13 @@ <h5><span class="header-section-number">7.1.0.1.1</span> Parametric vs. non-par
 ##  Shapiro-Wilk normality test
 ## 
 ## data:  sal_sub$length_2_mm
-## W = 0.93331, p-value &lt; 2.2e-16</code></pre>
+## W = 0.93584, p-value &lt; 2.2e-16</code></pre>
 <div class="sourceCode" id="cb151"><pre class="sourceCode r"><code class="sourceCode r"><span id="cb151-1"><a href="correlation-and-simple-linear-regression.html#cb151-1" tabindex="-1"></a><span class="fu">shapiro.test</span>(sal_sub<span class="sc">$</span>weight_g)</span></code></pre></div>
 <pre><code>## 
 ##  Shapiro-Wilk normality test
 ## 
 ## data:  sal_sub$weight_g
-## W = 0.56817, p-value &lt; 2.2e-16</code></pre>
+## W = 0.5666, p-value &lt; 2.2e-16</code></pre>
 <p>The <strong>null hypothesis of the Shapiro-Wilk normality test is that the variable is normally distributed</strong>, so a p-value less than 0.05, at 95% confidence (as we see for both of our variables here) tells use that <em>our data does not fit a normal distribution</em>.</p>
 <p>Therefore we have two options as we did with our t-test example in the previous lesson: transform the variables and see if that helps, or use a non-parametric test. Here, we will go ahead with the non-parametric Spearman correlation test.</p>
 <div class="alert alert-info">

docs/power.html

@@ -368,9 +368,9 @@ <h4><span class="header-section-number">9.0.1.1</span> Sample size<a href="power
 <span id="cb181-8"><a href="power.html#cb181-8" tabindex="-1"></a>  <span class="fu">ungroup</span>() <span class="sc">%&gt;%</span>   </span>
 <span id="cb181-9"><a href="power.html#cb181-9" tabindex="-1"></a>  <span class="fu">t_test</span>(bill_length_mm <span class="sc">~</span> species, <span class="at">var.equal =</span> <span class="cn">FALSE</span>, <span class="at">detailed =</span> <span class="cn">TRUE</span>)</span></code></pre></div>
 <pre><code>## # A tibble: 1 × 15
-##   estimate estimate1 estimate2 .y.           group1 group2    n1    n2 statistic       p    df conf.low
-## *    &lt;dbl&gt;     &lt;dbl&gt;     &lt;dbl&gt; &lt;chr&gt;         &lt;chr&gt;  &lt;chr&gt;  &lt;int&gt; &lt;int&gt;     &lt;dbl&gt;   &lt;dbl&gt; &lt;dbl&gt;    &lt;dbl&gt;
-## 1    -13.4      37.2      50.6 bill_length_… Adelie Gentoo     2     2     -15.7 0.00449  1.95    -17.2
+##   estimate estimate1 estimate2 .y.            group1 group2    n1    n2 statistic      p    df conf.low
+## *    &lt;dbl&gt;     &lt;dbl&gt;     &lt;dbl&gt; &lt;chr&gt;          &lt;chr&gt;  &lt;chr&gt;  &lt;int&gt; &lt;int&gt;     &lt;dbl&gt;  &lt;dbl&gt; &lt;dbl&gt;    &lt;dbl&gt;
+## 1    -8.65      39.6      48.3 bill_length_mm Adelie Gentoo     2     2     -4.20 0.0884  1.45    -21.7
 ## # ℹ 3 more variables: conf.high &lt;dbl&gt;, method &lt;chr&gt;, alternative &lt;chr&gt;</code></pre>
 <p>Compared to the p-value in the previous test, this p-value is much larger and, depending on which random samples were chosen in <code>slice_sample</code>, likely higher than our cut-off of 0.05 and would lead to an insignificant difference in bill length between the two species.</p>
 <p>Therefore, as sample size decreases, so does our power in identifying true trends.</p>
@@ -382,8 +382,8 @@ <h4><span class="header-section-number">9.0.1.2</span> Magnitude of effect<a hre
 <span id="cb183-2"><a href="power.html#cb183-2" tabindex="-1"></a>  <span class="fu">filter</span>(species <span class="sc">%in%</span> <span class="fu">c</span>(<span class="st">&quot;Adelie&quot;</span>, <span class="st">&quot;Gentoo&quot;</span>)) <span class="sc">%&gt;%</span></span>
 <span id="cb183-3"><a href="power.html#cb183-3" tabindex="-1"></a>  <span class="fu">ggplot</span>() <span class="sc">+</span></span>
 <span id="cb183-4"><a href="power.html#cb183-4" tabindex="-1"></a>  <span class="fu">geom_histogram</span>(<span class="fu">aes</span>(<span class="at">x =</span> bill_length_mm, <span class="at">fill =</span> species), <span class="at">alpha =</span> <span class="fl">0.50</span>) <span class="sc">+</span></span>
-<span id="cb183-5"><a href="power.html#cb183-5" tabindex="-1"></a>  <span class="fu">geom_vline</span>(<span class="at">xintercept =</span>  <span class="fl">38.8</span>) <span class="sc">+</span></span>
-<span id="cb183-6"><a href="power.html#cb183-6" tabindex="-1"></a>  <span class="fu">geom_vline</span>(<span class="at">xintercept =</span> <span class="fl">47.5</span>)</span></code></pre></div>
+<span id="cb183-5"><a href="power.html#cb183-5" tabindex="-1"></a>  <span class="fu">geom_vline</span>(<span class="at">xintercept =</span>  <span class="fl">38.8</span>, <span class="at">col =</span> <span class="st">&quot;black&quot;</span>) <span class="sc">+</span></span>
+<span id="cb183-6"><a href="power.html#cb183-6" tabindex="-1"></a>  <span class="fu">geom_vline</span>(<span class="at">xintercept =</span> <span class="fl">47.5</span>, <span class="at">col =</span> <span class="st">&quot;black&quot;</span>)</span></code></pre></div>
 <p><img src="_main_files/figure-html/unnamed-chunk-162-1.png" width="864" /></p>
 <p>In the grand scheme of things the difference between these two populations is relatively small in magnitude: their histograms overlap, indicating that some penguins in both species often have similar bill lengths. But, what if the bills of all of the Gentoo penguins <em>magically</em> grew an extra 15 mm?</p>
 <div class="sourceCode" id="cb184"><pre class="sourceCode r"><code class="sourceCode r"><span id="cb184-1"><a href="power.html#cb184-1" tabindex="-1"></a>penguins <span class="sc">%&gt;%</span></span>
@@ -392,8 +392,8 @@ <h4><span class="header-section-number">9.0.1.2</span> Magnitude of effect<a hre
 <span id="cb184-4"><a href="power.html#cb184-4" tabindex="-1"></a>  <span class="fu">mutate</span>(<span class="at">bill_length_mm =</span> <span class="fu">if_else</span>(species <span class="sc">==</span> <span class="st">&quot;Gentoo&quot;</span>, bill_length_mm <span class="sc">+</span> <span class="dv">15</span>, bill_length_mm)) <span class="sc">%&gt;%</span> </span>
 <span id="cb184-5"><a href="power.html#cb184-5" tabindex="-1"></a>  <span class="fu">ggplot</span>() <span class="sc">+</span> </span>
 <span id="cb184-6"><a href="power.html#cb184-6" tabindex="-1"></a>  <span class="fu">geom_histogram</span>(<span class="fu">aes</span>(<span class="at">x =</span> bill_length_mm, <span class="at">fill =</span> species), <span class="at">alpha =</span> <span class="fl">0.56</span>) <span class="sc">+</span></span>
-<span id="cb184-7"><a href="power.html#cb184-7" tabindex="-1"></a>  <span class="fu">geom_vline</span>(<span class="at">xintercept =</span>  <span class="fl">38.8</span>) <span class="sc">+</span></span>
-<span id="cb184-8"><a href="power.html#cb184-8" tabindex="-1"></a>  <span class="fu">geom_vline</span>(<span class="at">xintercept =</span> <span class="fl">62.5</span>) <span class="co"># I found this from the new mutated values separately</span></span></code></pre></div>
+<span id="cb184-7"><a href="power.html#cb184-7" tabindex="-1"></a>  <span class="fu">geom_vline</span>(<span class="at">xintercept =</span>  <span class="fl">38.8</span>, <span class="at">col =</span> <span class="st">&quot;black&quot;</span>) <span class="sc">+</span></span>
+<span id="cb184-8"><a href="power.html#cb184-8" tabindex="-1"></a>  <span class="fu">geom_vline</span>(<span class="at">xintercept =</span> <span class="fl">62.5</span>, <span class="at">col =</span> <span class="st">&quot;black&quot;</span>) <span class="co"># I found this from the new mutated values separately</span></span></code></pre></div>
 <p><img src="_main_files/figure-html/unnamed-chunk-163-1.png" width="864" /></p>
 <p>And now perform the t-test on this new <em>magical</em> data:</p>
 <div class="sourceCode" id="cb185"><pre class="sourceCode r"><code class="sourceCode r"><span id="cb185-1"><a href="power.html#cb185-1" tabindex="-1"></a>penguins <span class="sc">%&gt;%</span></span>
@@ -429,7 +429,7 @@ <h4><span class="header-section-number">9.0.1.2</span> Magnitude of effect<a hre
 <pre><code>## # A tibble: 1 × 15
 ##   estimate estimate1 estimate2 .y.            group1 group2    n1    n2 statistic      p    df conf.low
 ## *    &lt;dbl&gt;     &lt;dbl&gt;     &lt;dbl&gt; &lt;chr&gt;          &lt;chr&gt;  &lt;chr&gt;  &lt;int&gt; &lt;int&gt;     &lt;dbl&gt;  &lt;dbl&gt; &lt;dbl&gt;    &lt;dbl&gt;
-## 1    -24.8      39.1      64.0 bill_length_mm Adelie Gentoo     2     2     -6.88 0.0252  1.84    -41.7
+## 1    -19.8      40.4      60.1 bill_length_mm Adelie Gentoo     2     2     -15.2 0.0125  1.51    -27.5
 ## # ℹ 3 more variables: conf.high &lt;dbl&gt;, method &lt;chr&gt;, alternative &lt;chr&gt;</code></pre>
 </div>
 </div>