vega-regression - npm Package Compare versions

Comparing version 1.0.6 to 1.0.7

build/vega-regression.js

 (function (global, factory) {
   typeof exports === 'object' && typeof module !== 'undefined' ? factory(exports, require('vega-statistics'), require('vega-dataflow'), require('vega-util')) :
   typeof define === 'function' && define.amd ? define(['exports', 'vega-statistics', 'vega-dataflow', 'vega-util'], factory) :
-  (global = global || self, factory((global.vega = global.vega || {}, global.vega.transforms = {}), global.vega, global.vega, global.vega));
+  (global = typeof globalThis !== 'undefined' ? globalThis : global || self, factory((global.vega = global.vega || {}, global.vega.transforms = {}), global.vega, global.vega, global.vega));
 }(this, (function (exports, vegaStatistics, vegaDataflow, vegaUtil) { 'use strict';
@@ -57,34 +57,34 @@
 
-  var prototype = vegaUtil.inherits(Loess, vegaDataflow.Transform);
+  vegaUtil.inherits(Loess, vegaDataflow.Transform, {
+    transform(_, pulse) {
+      const out = pulse.fork(pulse.NO_SOURCE | pulse.NO_FIELDS);
 
-  prototype.transform = function(_, pulse) {
-    var out = pulse.fork(pulse.NO_SOURCE | pulse.NO_FIELDS);
+      if (!this.value || pulse.changed() || _.modified()) {
+        const source = pulse.materialize(pulse.SOURCE).source,
+              groups = partition(source, _.groupby),
+              names = (_.groupby || []).map(vegaUtil.accessorName),
+              m = names.length,
+              as = _.as || [vegaUtil.accessorName(_.x), vegaUtil.accessorName(_.y)],
+              values = [];
 
-    if (!this.value || pulse.changed() || _.modified()) {
-      const source = pulse.materialize(pulse.SOURCE).source,
-            groups = partition(source, _.groupby),
-            names = (_.groupby || []).map(vegaUtil.accessorName),
-            m = names.length,
-            as = _.as || [vegaUtil.accessorName(_.x), vegaUtil.accessorName(_.y)],
-            values = [];
-
-      groups.forEach(g => {
-        vegaStatistics.regressionLoess(g, _.x, _.y, _.bandwidth || 0.3).forEach(p => {
-          const t = {};
-          for (let i=0; i<m; ++i) {
-            t[names[i]] = g.dims[i];
-          }
-          t[as[0]] = p[0];
-          t[as[1]] = p[1];
-          values.push(vegaDataflow.ingest(t));
+        groups.forEach(g => {
+          vegaStatistics.regressionLoess(g, _.x, _.y, _.bandwidth || 0.3).forEach(p => {
+            const t = {};
+            for (let i=0; i<m; ++i) {
+              t[names[i]] = g.dims[i];
+            }
+            t[as[0]] = p[0];
+            t[as[1]] = p[1];
+            values.push(vegaDataflow.ingest(t));
+          });
         });
-      });
 
-      if (this.value) out.rem = this.value;
-      this.value = out.add = out.source = values;
+        if (this.value) out.rem = this.value;
+        this.value = out.add = out.source = values;
+      }
+
+      return out;
     }
+  });
 
-    return out;
-  };
-
   const Methods = {
@@ -99,5 +99,4 @@ linear: vegaStatistics.regressionLinear,
 
-  function degreesOfFreedom(method, order) {
-    return method === 'poly' ? order : method === 'quad' ? 2 : 1;
-  }
+  const degreesOfFreedom = (method, order) =>
+    method === 'poly' ? order : method === 'quad' ? 2 : 1;
 
@@ -134,77 +133,77 @@ /**
 
-  var prototype$1 = vegaUtil.inherits(Regression, vegaDataflow.Transform);
+  vegaUtil.inherits(Regression, vegaDataflow.Transform, {
+    transform(_, pulse) {
+      const out = pulse.fork(pulse.NO_SOURCE | pulse.NO_FIELDS);
 
-  prototype$1.transform = function(_, pulse) {
-    var out = pulse.fork(pulse.NO_SOURCE | pulse.NO_FIELDS);
+      if (!this.value || pulse.changed() || _.modified()) {
+        const source = pulse.materialize(pulse.SOURCE).source,
+              groups = partition(source, _.groupby),
+              names = (_.groupby || []).map(vegaUtil.accessorName),
+              method = _.method || 'linear',
+              order = _.order || 3,
+              dof = degreesOfFreedom(method, order),
+              as = _.as || [vegaUtil.accessorName(_.x), vegaUtil.accessorName(_.y)],
+              fit = Methods[method],
+              values = [];
 
-    if (!this.value || pulse.changed() || _.modified()) {
-      const source = pulse.materialize(pulse.SOURCE).source,
-            groups = partition(source, _.groupby),
-            names = (_.groupby || []).map(vegaUtil.accessorName),
-            method = _.method || 'linear',
-            order = _.order || 3,
-            dof = degreesOfFreedom(method, order),
-            as = _.as || [vegaUtil.accessorName(_.x), vegaUtil.accessorName(_.y)],
-            fit = Methods[method],
-            values = [];
+        let domain = _.extent;
 
-      let domain = _.extent;
-
-      if (!vegaUtil.hasOwnProperty(Methods, method)) {
-        vegaUtil.error('Invalid regression method: ' + method);
-      }
-
-      if (domain != null) {
-        if (method === 'log' && domain[0] <= 0) {
-          pulse.dataflow.warn('Ignoring extent with values <= 0 for log regression.');
-          domain = null;
+        if (!vegaUtil.hasOwnProperty(Methods, method)) {
+          vegaUtil.error('Invalid regression method: ' + method);
         }
-      }
 
-      groups.forEach(g => {
-        const n = g.length;
-        if (n <= dof) {
-          pulse.dataflow.warn('Skipping regression with more parameters than data points.');
-          return;
+        if (domain != null) {
+          if (method === 'log' && domain[0] <= 0) {
+            pulse.dataflow.warn('Ignoring extent with values <= 0 for log regression.');
+            domain = null;
+          }
         }
 
-        const model = fit(g, _.x, _.y, order);
+        groups.forEach(g => {
+          const n = g.length;
+          if (n <= dof) {
+            pulse.dataflow.warn('Skipping regression with more parameters than data points.');
+            return;
+          }
 
-        if (_.params) {
-          // if parameter vectors requested return those
-          values.push(vegaDataflow.ingest({
-            keys: g.dims,
-            coef: model.coef,
-            rSquared: model.rSquared
-          }));
-          return;
-        }
+          const model = fit(g, _.x, _.y, order);
 
-        const dom = domain || vegaUtil.extent(g, _.x),
-              add = p => {
-                const t = {};
-                for (let i=0; i<names.length; ++i) {
-                  t[names[i]] = g.dims[i];
-                }
-                t[as[0]] = p[0];
-                t[as[1]] = p[1];
-                values.push(vegaDataflow.ingest(t));
-              };
+          if (_.params) {
+            // if parameter vectors requested return those
+            values.push(vegaDataflow.ingest({
+              keys: g.dims,
+              coef: model.coef,
+              rSquared: model.rSquared
+            }));
+            return;
+          }
 
-        if (method === 'linear') {
-          // for linear regression we only need the end points
-          dom.forEach(x => add([x, model.predict(x)]));
-        } else {
-          // otherwise return trend line sample points
-          vegaStatistics.sampleCurve(model.predict, dom, 25, 200).forEach(add);
-        }
-      });
+          const dom = domain || vegaUtil.extent(g, _.x),
+                add = p => {
+                  const t = {};
+                  for (let i=0; i<names.length; ++i) {
+                    t[names[i]] = g.dims[i];
+                  }
+                  t[as[0]] = p[0];
+                  t[as[1]] = p[1];
+                  values.push(vegaDataflow.ingest(t));
+                };
 
-      if (this.value) out.rem = this.value;
-      this.value = out.add = out.source = values;
+          if (method === 'linear') {
+            // for linear regression we only need the end points
+            dom.forEach(x => add([x, model.predict(x)]));
+          } else {
+            // otherwise return trend line sample points
+            vegaStatistics.sampleCurve(model.predict, dom, 25, 200).forEach(add);
+          }
+        });
+
+        if (this.value) out.rem = this.value;
+        this.value = out.add = out.source = values;
+      }
+
+      return out;
     }
+  });
 
-    return out;
-  };
-
   exports.loess = Loess;
@@ -211,0 +210,0 @@ exports.regression = Regression;

build/vega-regression.min.js

@@ -1,1 +0,1 @@
-!function(e,r){"object"==typeof exports&&"undefined"!=typeof module?r(exports,require("vega-statistics"),require("vega-dataflow"),require("vega-util")):"function"==typeof define&&define.amd?define(["exports","vega-statistics","vega-dataflow","vega-util"],r):r(((e=e||self).vega=e.vega||{},e.vega.transforms={}),e.vega,e.vega,e.vega)}(this,(function(e,r,a,t){"use strict";function n(e,r){var a,t,n,s,i,o,u=[],l=function(e){return e(s)};if(null==r)u.push(e);else for(a={},t=0,n=e.length;t<n;++t)s=e[t],(o=a[i=r.map(l)])||(a[i]=o=[],o.dims=i,u.push(o)),o.push(s);return u}function s(e){a.Transform.call(this,null,e)}s.Definition={type:"Loess",metadata:{generates:!0},params:[{name:"x",type:"field",required:!0},{name:"y",type:"field",required:!0},{name:"groupby",type:"field",array:!0},{name:"bandwidth",type:"number",default:.3},{name:"as",type:"string",array:!0}]},t.inherits(s,a.Transform).transform=function(e,s){var i=s.fork(s.NO_SOURCE|s.NO_FIELDS);if(!this.value||s.changed()||e.modified()){const o=n(s.materialize(s.SOURCE).source,e.groupby),u=(e.groupby||[]).map(t.accessorName),l=u.length,d=e.as||[t.accessorName(e.x),t.accessorName(e.y)],f=[];o.forEach(t=>{r.regressionLoess(t,e.x,e.y,e.bandwidth||.3).forEach(e=>{const r={};for(let e=0;e<l;++e)r[u[e]]=t.dims[e];r[d[0]]=e[0],r[d[1]]=e[1],f.push(a.ingest(r))})}),this.value&&(i.rem=this.value),this.value=i.add=i.source=f}return i};const i={linear:r.regressionLinear,log:r.regressionLog,exp:r.regressionExp,pow:r.regressionPow,quad:r.regressionQuad,poly:r.regressionPoly};function o(e){a.Transform.call(this,null,e)}o.Definition={type:"Regression",metadata:{generates:!0},params:[{name:"x",type:"field",required:!0},{name:"y",type:"field",required:!0},{name:"groupby",type:"field",array:!0},{name:"method",type:"string",default:"linear",values:Object.keys(i)},{name:"order",type:"number",default:3},{name:"extent",type:"number",array:!0,length:2},{name:"params",type:"boolean",default:!1},{name:"as",type:"string",array:!0}]},t.inherits(o,a.Transform).transform=function(e,s){var o=s.fork(s.NO_SOURCE|s.NO_FIELDS);if(!this.value||s.changed()||e.modified()){const u=n(s.materialize(s.SOURCE).source,e.groupby),l=(e.groupby||[]).map(t.accessorName),d=e.method||"linear",f=e.order||3,m=function(e,r){return"poly"===e?r:"quad"===e?2:1}(d,f),p=e.as||[t.accessorName(e.x),t.accessorName(e.y)],c=i[d],g=[];let y=e.extent;t.hasOwnProperty(i,d)||t.error("Invalid regression method: "+d),null!=y&&"log"===d&&y[0]<=0&&(s.dataflow.warn("Ignoring extent with values <= 0 for log regression."),y=null),u.forEach(n=>{if(n.length<=m)return void s.dataflow.warn("Skipping regression with more parameters than data points.");const i=c(n,e.x,e.y,f);if(e.params)return void g.push(a.ingest({keys:n.dims,coef:i.coef,rSquared:i.rSquared}));const o=y||t.extent(n,e.x),u=e=>{const r={};for(let e=0;e<l.length;++e)r[l[e]]=n.dims[e];r[p[0]]=e[0],r[p[1]]=e[1],g.push(a.ingest(r))};"linear"===d?o.forEach(e=>u([e,i.predict(e)])):r.sampleCurve(i.predict,o,25,200).forEach(u)}),this.value&&(o.rem=this.value),this.value=o.add=o.source=g}return o},e.loess=s,e.regression=o,Object.defineProperty(e,"__esModule",{value:!0})}));
+!function(e,r){"object"==typeof exports&&"undefined"!=typeof module?r(exports,require("vega-statistics"),require("vega-dataflow"),require("vega-util")):"function"==typeof define&&define.amd?define(["exports","vega-statistics","vega-dataflow","vega-util"],r):r(((e="undefined"!=typeof globalThis?globalThis:e||self).vega=e.vega||{},e.vega.transforms={}),e.vega,e.vega,e.vega)}(this,(function(e,r,a,t){"use strict";function s(e,r){var a,t,s,n,o,i,l=[],u=function(e){return e(n)};if(null==r)l.push(e);else for(a={},t=0,s=e.length;t<s;++t)n=e[t],(i=a[o=r.map(u)])||(a[o]=i=[],i.dims=o,l.push(i)),i.push(n);return l}function n(e){a.Transform.call(this,null,e)}n.Definition={type:"Loess",metadata:{generates:!0},params:[{name:"x",type:"field",required:!0},{name:"y",type:"field",required:!0},{name:"groupby",type:"field",array:!0},{name:"bandwidth",type:"number",default:.3},{name:"as",type:"string",array:!0}]},t.inherits(n,a.Transform,{transform(e,n){const o=n.fork(n.NO_SOURCE|n.NO_FIELDS);if(!this.value||n.changed()||e.modified()){const i=s(n.materialize(n.SOURCE).source,e.groupby),l=(e.groupby||[]).map(t.accessorName),u=l.length,d=e.as||[t.accessorName(e.x),t.accessorName(e.y)],f=[];i.forEach(t=>{r.regressionLoess(t,e.x,e.y,e.bandwidth||.3).forEach(e=>{const r={};for(let e=0;e<u;++e)r[l[e]]=t.dims[e];r[d[0]]=e[0],r[d[1]]=e[1],f.push(a.ingest(r))})}),this.value&&(o.rem=this.value),this.value=o.add=o.source=f}return o}});const o={linear:r.regressionLinear,log:r.regressionLog,exp:r.regressionExp,pow:r.regressionPow,quad:r.regressionQuad,poly:r.regressionPoly};function i(e){a.Transform.call(this,null,e)}i.Definition={type:"Regression",metadata:{generates:!0},params:[{name:"x",type:"field",required:!0},{name:"y",type:"field",required:!0},{name:"groupby",type:"field",array:!0},{name:"method",type:"string",default:"linear",values:Object.keys(o)},{name:"order",type:"number",default:3},{name:"extent",type:"number",array:!0,length:2},{name:"params",type:"boolean",default:!1},{name:"as",type:"string",array:!0}]},t.inherits(i,a.Transform,{transform(e,n){const i=n.fork(n.NO_SOURCE|n.NO_FIELDS);if(!this.value||n.changed()||e.modified()){const l=s(n.materialize(n.SOURCE).source,e.groupby),u=(e.groupby||[]).map(t.accessorName),d=e.method||"linear",f=e.order||3,m=((e,r)=>"poly"===e?r:"quad"===e?2:1)(d,f),p=e.as||[t.accessorName(e.x),t.accessorName(e.y)],g=o[d],c=[];let y=e.extent;t.hasOwnProperty(o,d)||t.error("Invalid regression method: "+d),null!=y&&"log"===d&&y[0]<=0&&(n.dataflow.warn("Ignoring extent with values <= 0 for log regression."),y=null),l.forEach(s=>{if(s.length<=m)return void n.dataflow.warn("Skipping regression with more parameters than data points.");const o=g(s,e.x,e.y,f);if(e.params)return void c.push(a.ingest({keys:s.dims,coef:o.coef,rSquared:o.rSquared}));const i=y||t.extent(s,e.x),l=e=>{const r={};for(let e=0;e<u.length;++e)r[u[e]]=s.dims[e];r[p[0]]=e[0],r[p[1]]=e[1],c.push(a.ingest(r))};"linear"===d?i.forEach(e=>l([e,o.predict(e)])):r.sampleCurve(o.predict,i,25,200).forEach(l)}),this.value&&(i.rem=this.value),this.value=i.add=i.source=c}return i}}),e.loess=n,e.regression=i,Object.defineProperty(e,"__esModule",{value:!0})}));

package.json

 {
   "name": "vega-regression",
-  "version": "1.0.6",
+  "version": "1.0.7",
   "description": "Regression transform for Vega dataflows.",
@@ -26,6 +26,6 @@ "keywords": [
   "dependencies": {
-    "d3-array": "^2.4.0",
-    "vega-dataflow": "^5.5.1",
-    "vega-statistics": "^1.7.4",
-    "vega-util": "^1.13.2"
+    "d3-array": "^2.5.1",
+    "vega-dataflow": "^5.7.1",
+    "vega-statistics": "^1.7.7",
+    "vega-util": "^1.15.0"
   },
@@ -35,3 +35,3 @@ "devDependencies": {
   },
-  "gitHead": "35e31c5c6b54db9dc3a577b5adad8d15ec274d32"
+  "gitHead": "28db83352e43e321dfe55fc5cb6489b211e45662"
 }

src/Loess.js

@@ -31,32 +31,32 @@ import partition from './partition';
 
-var prototype = inherits(Loess, Transform);
+inherits(Loess, Transform, {
+  transform(_, pulse) {
+    const out = pulse.fork(pulse.NO_SOURCE | pulse.NO_FIELDS);
 
-prototype.transform = function(_, pulse) {
-  var out = pulse.fork(pulse.NO_SOURCE | pulse.NO_FIELDS);
+    if (!this.value || pulse.changed() || _.modified()) {
+      const source = pulse.materialize(pulse.SOURCE).source,
+            groups = partition(source, _.groupby),
+            names = (_.groupby || []).map(accessorName),
+            m = names.length,
+            as = _.as || [accessorName(_.x), accessorName(_.y)],
+            values = [];
 
-  if (!this.value || pulse.changed() || _.modified()) {
-    const source = pulse.materialize(pulse.SOURCE).source,
-          groups = partition(source, _.groupby),
-          names = (_.groupby || []).map(accessorName),
-          m = names.length,
-          as = _.as || [accessorName(_.x), accessorName(_.y)],
-          values = [];
-
-    groups.forEach(g => {
-      regressionLoess(g, _.x, _.y, _.bandwidth || 0.3).forEach(p => {
-        const t = {};
-        for (let i=0; i<m; ++i) {
-          t[names[i]] = g.dims[i];
-        }
-        t[as[0]] = p[0];
-        t[as[1]] = p[1];
-        values.push(ingest(t));
+      groups.forEach(g => {
+        regressionLoess(g, _.x, _.y, _.bandwidth || 0.3).forEach(p => {
+          const t = {};
+          for (let i=0; i<m; ++i) {
+            t[names[i]] = g.dims[i];
+          }
+          t[as[0]] = p[0];
+          t[as[1]] = p[1];
+          values.push(ingest(t));
+        });
       });
-    });
 
-    if (this.value) out.rem = this.value;
-    this.value = out.add = out.source = values;
+      if (this.value) out.rem = this.value;
+      this.value = out.add = out.source = values;
+    }
+
+    return out;
   }
-
-  return out;
-};
+});

src/Regression.js

@@ -18,5 +18,4 @@ import partition from './partition';
 
-function degreesOfFreedom(method, order) {
-  return method === 'poly' ? order : method === 'quad' ? 2 : 1;
-}
+const degreesOfFreedom = (method, order) =>
+  method === 'poly' ? order : method === 'quad' ? 2 : 1;
 
@@ -53,75 +52,75 @@ /**
 
-var prototype = inherits(Regression, Transform);
+inherits(Regression, Transform, {
+  transform(_, pulse) {
+    const out = pulse.fork(pulse.NO_SOURCE | pulse.NO_FIELDS);
 
-prototype.transform = function(_, pulse) {
-  var out = pulse.fork(pulse.NO_SOURCE | pulse.NO_FIELDS);
+    if (!this.value || pulse.changed() || _.modified()) {
+      const source = pulse.materialize(pulse.SOURCE).source,
+            groups = partition(source, _.groupby),
+            names = (_.groupby || []).map(accessorName),
+            method = _.method || 'linear',
+            order = _.order || 3,
+            dof = degreesOfFreedom(method, order),
+            as = _.as || [accessorName(_.x), accessorName(_.y)],
+            fit = Methods[method],
+            values = [];
 
-  if (!this.value || pulse.changed() || _.modified()) {
-    const source = pulse.materialize(pulse.SOURCE).source,
-          groups = partition(source, _.groupby),
-          names = (_.groupby || []).map(accessorName),
-          method = _.method || 'linear',
-          order = _.order || 3,
-          dof = degreesOfFreedom(method, order),
-          as = _.as || [accessorName(_.x), accessorName(_.y)],
-          fit = Methods[method],
-          values = [];
+      let domain = _.extent;
 
-    let domain = _.extent;
-
-    if (!hasOwnProperty(Methods, method)) {
-      error('Invalid regression method: ' + method);
-    }
-
-    if (domain != null) {
-      if (method === 'log' && domain[0] <= 0) {
-        pulse.dataflow.warn('Ignoring extent with values <= 0 for log regression.');
-        domain = null;
+      if (!hasOwnProperty(Methods, method)) {
+        error('Invalid regression method: ' + method);
       }
-    }
 
-    groups.forEach(g => {
-      const n = g.length;
-      if (n <= dof) {
-        pulse.dataflow.warn('Skipping regression with more parameters than data points.');
-        return;
+      if (domain != null) {
+        if (method === 'log' && domain[0] <= 0) {
+          pulse.dataflow.warn('Ignoring extent with values <= 0 for log regression.');
+          domain = null;
+        }
       }
 
-      const model = fit(g, _.x, _.y, order);
+      groups.forEach(g => {
+        const n = g.length;
+        if (n <= dof) {
+          pulse.dataflow.warn('Skipping regression with more parameters than data points.');
+          return;
+        }
 
-      if (_.params) {
-        // if parameter vectors requested return those
-        values.push(ingest({
-          keys: g.dims,
-          coef: model.coef,
-          rSquared: model.rSquared
-        }));
-        return;
-      }
+        const model = fit(g, _.x, _.y, order);
 
-      const dom = domain || extent(g, _.x),
-            add = p => {
-              const t = {};
-              for (let i=0; i<names.length; ++i) {
-                t[names[i]] = g.dims[i];
-              }
-              t[as[0]] = p[0];
-              t[as[1]] = p[1];
-              values.push(ingest(t));
-            };
+        if (_.params) {
+          // if parameter vectors requested return those
+          values.push(ingest({
+            keys: g.dims,
+            coef: model.coef,
+            rSquared: model.rSquared
+          }));
+          return;
+        }
 
-      if (method === 'linear') {
-        // for linear regression we only need the end points
-        dom.forEach(x => add([x, model.predict(x)]));
-      } else {
-        // otherwise return trend line sample points
-        sampleCurve(model.predict, dom, 25, 200).forEach(add);
-      }
-    });
+        const dom = domain || extent(g, _.x),
+              add = p => {
+                const t = {};
+                for (let i=0; i<names.length; ++i) {
+                  t[names[i]] = g.dims[i];
+                }
+                t[as[0]] = p[0];
+                t[as[1]] = p[1];
+                values.push(ingest(t));
+              };
 
-    if (this.value) out.rem = this.value;
-    this.value = out.add = out.source = values;
+        if (method === 'linear') {
+          // for linear regression we only need the end points
+          dom.forEach(x => add([x, model.predict(x)]));
+        } else {
+          // otherwise return trend line sample points
+          sampleCurve(model.predict, dom, 25, 200).forEach(add);
+        }
+      });
+
+      if (this.value) out.rem = this.value;
+      this.value = out.add = out.source = values;
+    }
+
+    return out;
   }
-
-  return out;
-};
+});

New alerts

License Policy Violation

License

This package is not allowed per your license policy. Review the package's license to ensure compliance.

Found 1 instance in 1 package

Fixed alerts

License Policy Violation

License

This package is not allowed per your license policy. Review the package's license to ensure compliance.

Found 1 instance in 1 package

Improved metrics

Total package byte prevSize

20753

increased by1.3%

Lines of code

368

increased by0.55%

Dependency changes

• Updatedd3-array@^2.5.1

• Updatedvega-dataflow@^5.7.1

• Updatedvega-statistics@^1.7.7

• Updatedvega-util@^1.15.0