simplify

.gitignore

@@ -17,3 +17,4 @@ __pycache__/
 /vignettes/*.html
 /vignettes/*.R
 /vignettes/*_files/
+/vignettes/.build.timestamp

.vscode/c_cpp_properties.json

@@ -11,6 +11,7 @@
             "cStandard": "c11",
             "cppStandard": "c++11",
             "intelliSenseMode": "linux-clang-x64",
+            "compilerArgs": ["-Wall"]
         },
     ],
     "version": 4,

.vscode/settings.json

@@ -31,4 +31,20 @@
         "ylen",
         "zlen"
     ],
+    "files.associations": {
+        "*.desktop": "ini",
+        "vimrc": "viml",
+        "zshrc": "shellscript",
+        ".zpreztorc": "shellscript",
+        "zpreztorc": "shellscript",
+        "asv.conf.json": "jsonc",
+        "*.tags": "lua",
+        "*.tikz": "context",
+        "*.lmtx": "context",
+        ".prettierrc": "jsonc",
+        "*.yaml": "jinja-yaml",
+        "*.j2": "jinja-md",
+        "cmath": "cpp",
+        "cstddef": "cpp"
+    },
 }

R/censoring.r

@@ -36,17 +36,17 @@ validate_censoring <- function(data, sigma, dists, censor_val, censor_range, mis
 		stop('You have to provide both a censoring value and a censor_range or none')
 
 	if (!is.null(censor_range)) {
-		if (!is.numeric(censor_val) || (length(censor_val) != 1L && length(censor_val) != g))
-		stop('censor_val has to be a single numeric value, or length(censor_val) == ncol(data) must be TRUE')
+		if (!is.numeric(censor_val) || !(length(censor_val) %in% c(g, 1L)))
+			stop('censor_val has to be a single numeric value, or length(censor_val) == ncol(data) must be TRUE')
 
 		if (!is.numeric(censor_range) || !(nrow(censor_range) %in% c(g, 1L)) || ncol(censor_range) != 2L || any(diff(t(censor_range)) <= 0L))
-		stop('censor_range has to be a numeric vector of length 2, the second of which being larger,
+			stop('censor_range has to be a numeric vector of length 2, the second of which being larger,
 				 or a matrix with nrow(censor_range) == ncol(data) where each row is such a vector')
 	}
 
 	if (!is.null(missing_range)) {
 		if (!is.numeric(missing_range) || !(nrow(missing_range) %in% c(g, 1L)) || ncol(missing_range) != 2L || any(diff(t(missing_range)) <= 0L))
-		stop('missing_range has to be a numeric vector of length 2, the second of which being larger,
+			stop('missing_range has to be a numeric vector of length 2, the second of which being larger,
 				 or a matrix with nrow(missing_range) == ncol(data) where each row is such a vector')
 	}
 

src/censoring.cpp

@@ -18,9 +18,7 @@ inline double censor_pair(
 	bool use_d;
 	const bool
 		one_uncertain = (c == thr) || (d == thr),
-		one_missing = NumericVector::is_na(c) || NumericVector::is_na(d),
-		one_of_each_invalid = one_uncertain && one_missing,
-		both_valid = !one_uncertain && !one_missing;
+		one_missing = NumericVector::is_na(c) || NumericVector::is_na(d);
 
 	/*
 	Rcout << "one uncertain=" << one_uncertain << " one missing=" << one_missing
@@ -45,36 +43,34 @@ inline double censor_pair(
 	// 3. (missing-uncertain): interference of two different box functions
 	// 4. (invalid×2): two missing or two uncertain => 1
 
-	if (both_valid) { //1
+	if (!one_uncertain && !one_missing) { //1
 		return exp(-pow(c-d, 2) / (kt*2));
-	} else { //at least one invalid
-		if (one_of_each_invalid) { //3
-			const double uncertain_range = uncertain1 - uncertain0 + 2*sigma;
-			return uncertain_range / (sqrt(uncertain_range) * sqrt(missing1 - missing0));
-		} else if (one_uncertain || one_missing) { //2
-			double m0, m1;
-			if (one_uncertain && !one_missing) {
-				m0  = uncertain0;
-				m1  = uncertain1;
-				use_d = c == thr;
-			} else if (!one_uncertain && one_missing) {
-				m0  = missing0;
-				m1  = missing1;
-				use_d = NumericVector::is_na(c);
-			}
-
-			const double v = use_d ? d : c;
-
-			return
-				pow(M_PI*kt/2, -1./4)
-				* sqrt(M_PI*kt/4)
-				* ( std::erfc((m0-v) / sigma) - std::erfc((m1-v) / sigma) )
-				/ sqrt(m1-m0);
-
-			//Rcout << "m0=" << m0 << " m1=" << m1 << " use d=" << use_d << " x=" << x << " v=" << v << '\n';
-		} else {// both are uncertain or both are missing => x *= 1 => noop
-			return 1;
+	} else if (one_uncertain && one_missing) { //3
+		const double uncertain_range = uncertain1 - uncertain0 + 2*sigma;
+		return uncertain_range / (sqrt(uncertain_range) * sqrt(missing1 - missing0));
+	} else if (one_uncertain || one_missing) { //2
+		double m0, m1;
+		if (one_uncertain && !one_missing) {
+			m0  = uncertain0;
+			m1  = uncertain1;
+			use_d = c == thr;
+		} else if (!one_uncertain && one_missing) {
+			m0  = missing0;
+			m1  = missing1;
+			use_d = NumericVector::is_na(c);
 		}
+
+		const double v = use_d ? d : c;
+
+		return
+			pow(M_PI*kt/2, -1./4)
+			* sqrt(M_PI*kt/4)
+			* ( std::erfc((m0-v) / sigma) - std::erfc((m1-v) / sigma) )
+			/ sqrt(m1-m0);
+
+		//Rcout << "m0=" << m0 << " m1=" << m1 << " use d=" << use_d << " x=" << x << " v=" << v << '\n';
+	} else {// both are uncertain or both are missing => x *= 1 => noop
+		return 1;
 	}
 }
 

vignettes/Diffusion-Maps.Rmd

@@ -168,16 +168,15 @@ plot3d(
     type = 's', radius = .01)
 view3d(theta = 10, phi = 30, zoom = .8)
 # now use your mouse to rotate the plot in the window
-rgl.close()
+close3d()
 ```
 
 For the popular `ggplot2` package, there is built in support in the form of a `fortify.DiffusionMap` method, which allows to use DiffusionMap objects as `data` parameter in the `ggplot` and `qplot` functions. But the regular `plot` function already returns a `ggplot` object anyway.
 
 ```{r}
 library(ggplot2)
-qplot(DC1, DC2, data = dm, colour = factor(num_cells)) +
-    scale_color_cube_helix()
-# or alternatively: ggplot(dif, aes(DC1, DC2, colour = factor(num.cells))) + ...
+ggplot(dm, aes(DC1, DC2, colour = factor(num_cells))) +
+    geom_point() + scale_color_cube_helix()
 ```
 
 As aesthetics, all diffusion components, gene expressions, and annotations are available. If you plan to make many plots, create a `data.frame` first by using `as.data.frame(dif)` or `fortify(dif)`, assign it to a variable name, and use it for plotting.
@@ -195,8 +194,8 @@ Dimensions `dims`
 Diffusion maps consist of the eigenvectors (which we refer to as diffusion components) and corresponding eigenvalues of the diffusion distance matrix. The latter indicate the diffusion components’ importance, i.e. how well the eigenvectors approximate the data. The eigenvectors are decreasingly meaningful.
 
 ```{r}
-qplot(y = eigenvalues(dm)) + theme_minimal() +
-    labs(x = 'Diffusion component (DC)', y = 'Eigenvalue')
+ggplot(NULL, aes(x = seq_along(eigenvalues(dm)), y = eigenvalues(dm))) + theme_minimal() +
+    geom_point() + labs(x = 'Diffusion component (DC)', y = 'Eigenvalue')
 ```
 
 The later DCs often become noisy and less useful: