SOCR Data PD BiomedBigMetadata - Revision history

Dinov: /* R-Script for Data Analytics */

2016-04-12T13:15:20Z

‎R-Script for Data Analytics

Dinov: /* Converting Dx to a dichotomous variable */

2016-04-12T13:13:29Z

‎Converting Dx to a dichotomous variable

Dinov: /* Automated loading of the PD data from Wiki Page */

2015-12-05T21:30:44Z

‎Automated loading of the PD data from Wiki Page

Dinov: /* More realistic simulation with non-trivial correlations */

2015-12-03T16:50:30Z

‎More realistic simulation with non-trivial correlations

Dinov: /* Automated loading of the PD data from Wiki Page */

2015-12-03T16:46:42Z

‎Automated loading of the PD data from Wiki Page

Dinov: /* More realistic simulation with non-trivial correlations */

2015-12-03T16:32:57Z

‎More realistic simulation with non-trivial correlations

Dinov at 15:42, 3 December 2015

2015-12-03T15:42:32Z

Dinov: Created page with "== SOCR Data - Predictive Big Data Analytics, Modeling, Analysis and Visualization of Clinical, Genetic and Imaging Data for Parkinson’s Disease == ==Summary== This cas..."

2015-09-02T18:59:55Z

Created page with "== SOCR Data - Predictive Big Data Analytics, Modeling, Analysis and Visualization of Clinical, Genetic and Imaging Data for Parkinson’s Disease == ==Summary== This cas..."

Show changes

← Older revision		Revision as of 13:15, 12 April 2016
Line 2,304:		Line 2,304:

	===Analytics===		===Analytics===
		+
		+	====Predictive Analytics Example using the PD Dataset====
		+
		+	[[SMHS_Usage_Rfundamentals#Example:_Generate_an_R_HTML_Export_using_the_SOCR_PD_Data\| This SOCR Activity illustrates predictive big data analytics using the Parkinson's dataset]].

	==== Converting Dx to a dichotomous variable====		==== Converting Dx to a dichotomous variable====

@@ Line 2,306: / Line 2,306: @@
 ==== Converting Dx to a dichotomous variable====
-  pd_data$Dx <- gsub("PD", 1, pd_data$Dx)
+  pd_data$\$$Dx <- gsub("PD", 1, pd_data$\$$Dx)
-  pd_data$Dx <- gsub("HC", 0, pd_data$Dx)
+  pd_data$\$$Dx <- gsub("HC", 0, pd_data$\$$Dx)
-  pd_data$Dx <- gsub("SWEDD", 0, pd_data$Dx)
+  pd_data$\$$Dx <- gsub("SWEDD", 0, pd_data$\$$Dx)
-  pd_data$Dx <- as.numeric(pd_data$Dx)
+  pd_data$\$$Dx <- as.numeric(pd_data$\$$Dx)
 ====Regression Analysis====

@@ Line 2,300: / Line 2,300: @@
   html_nodes(wiki_url, "#content")
-  pd_data <- html_table(html_nodes(wiki_url,"table")[[1]])
+  pd_data <- html_table(html_nodes(wiki_url,"table")\([[1]]\))
   head(pd_data); summary(pd_data)
-−
 ===Analytics===

@@ Line 2,518: / Line 2,518: @@
   # W = [Cholesky (R)][X]
-  # R = matrix(cbind(1,.80,.2,0.1,    .80,1,.7,.3,  .2,.7,1,.4,   .1,.3,.4,1),nrow=4);R
+  R = matrix(cbind(1,.80,.2,0.1,    .80,1,.7,.3,  .2,.7,1,.4,   .1,.3,.4,1),nrow=4);R
-  # U = t(chol(R))
+  U = t(chol(R))
-  # nvars = dim(U)[1]; numobs = 100000; set.seed(1234)
+  nvars = dim(U)[1]; numobs = 100000; set.seed(1234)
-  # random.normal = matrix(rnorm(nvars*numobs,0,1), nrow=nvars, ncol=numobs);
+  random.normal = matrix(rnorm(nvars*numobs,0,1), nrow=nvars, ncol=numobs);
-  # X = U %*% random.normal; newX = t(X); raw = as.data.frame(newX)
+  X = U %*% random.normal; newX = t(X); raw = as.data.frame(newX)
-  # orig.raw = as.data.frame(t(random.normal))
+  orig.raw = as.data.frame(t(random.normal))
-  # names(raw) = c("response","predictor1","predictor2","predictor3")
+  names(raw) = c("response","predictor1","predictor2","predictor3")
-  # cor(raw); plot(head(raw, 100)); plot(head(orig.raw,100))
+  cor(raw); plot(head(raw, 100)); plot(head(orig.raw,100))
 ==See also==

@@ Line 2,294: / Line 2,294: @@
 ==R-Script for Data Analytics ==
 ===Automated loading of the PD data from Wiki Page===
 ===R-Script for simulating PD data===

@@ Line 2,451: / Line 2,451: @@
   # For a specific correlation matrix, multiplying the Cholesky decomposition of the correlation
   # matrix by the data matrix generates a transformed dataset with the specified correlation.
-  # W = \left[Cholesky (R)\right]\left[X\right]
+  # W = [Cholesky (R)][X]
   # R = matrix(cbind(1,.80,.2,0.1,    .80,1,.7,.3,  .2,.7,1,.4,   .1,.3,.4,1),nrow=4);R

@@ Line 2,292: / Line 2,292: @@
 </center>
-==R-Script for simulating PD data==
+==R-Script for Data Analytics ==
   # Define number of subjects
   NumSubj <- 282
@@ Line 2,442: / Line 2,445: @@
   # Write out (save) the result to a file that can be shared
   write.table(sim_PD_Data, "output_data.csv", sep=",", row.names=FALSE, col.names=TRUE)
 ==See also==