improve naming and update to core and etable changes

actrf/actrf.go

@@ -102,14 +102,14 @@ func (af *RF) Add(act, src etensor.Tensor, thr float32) {
 	aNy, aNx, sNy, sNx := shp[0], shp[1], shp[2], shp[3]
 	for sy := 0; sy < sNy; sy++ {
 		for sx := 0; sx < sNx; sx++ {
-			tv := float32(etensor.Prjn2DVal(src, false, sy, sx))
+			tv := float32(etensor.Prjn2DValue(src, false, sy, sx))
 			if tv < thr {
 				continue
 			}
 			af.SumSrc.AddScalar([]int{sy, sx}, tv)
 			for ay := 0; ay < aNy; ay++ {
 				for ax := 0; ax < aNx; ax++ {
-					av := float32(etensor.Prjn2DVal(act, false, ay, ax))
+					av := float32(etensor.Prjn2DValue(act, false, ay, ax))
 					af.SumProd.AddScalar([]int{ay, ax, sy, sx}, av*tv)
 				}
 			}

actrf/running.go

@@ -20,10 +20,10 @@ func RunningAvg(out *etensor.Float32, act, src etensor.Tensor, tau float32) {
 	out.SetShape(oshp, nil, []string{"ActY", "ActX", "SrcY", "SrcX"})
 	for ay := 0; ay < aNy; ay++ {
 		for ax := 0; ax < aNx; ax++ {
-			av := float32(etensor.Prjn2DVal(act, false, ay, ax))
+			av := float32(etensor.Prjn2DValue(act, false, ay, ax))
 			for ty := 0; ty < tNy; ty++ {
 				for tx := 0; tx < tNx; tx++ {
-					tv := float32(etensor.Prjn2DVal(src, false, ty, tx))
+					tv := float32(etensor.Prjn2DValue(src, false, ty, tx))
 					oi := []int{ay, ax, ty, tx}
 					oo := out.Offset(oi)
 					ov := out.Values[oo]

confusion/confusion.go

@@ -126,16 +126,16 @@ func (cm *Matrix) SumTFPN(class int) {
 	for c := 0; c < n; c++ {
 		for r := 0; r < n; r++ {
 			if r == class && c == class { //        True Positive
-				v := cm.Sum.FloatValRowCell(r, c)
+				v := cm.Sum.FloatValueRowCell(r, c)
 				cm.TFPN.SetFloatRowCell(class, 0, v)
 			} else if r == class && c != class { // False Positive
-				fn += cm.Sum.FloatValRowCell(r, c)
+				fn += cm.Sum.FloatValueRowCell(r, c)
 				cm.TFPN.SetFloatRowCell(class, 1, fp)
 			} else if r != class && c == class { // False Negative
-				fp += cm.Sum.FloatValRowCell(r, c)
+				fp += cm.Sum.FloatValueRowCell(r, c)
 				cm.TFPN.SetFloatRowCell(class, 2, fn)
 			} else { //                             True Negative
-				tn += cm.Sum.FloatValRowCell(r, c)
+				tn += cm.Sum.FloatValueRowCell(r, c)
 				cm.TFPN.SetFloatRowCell(class, 3, tn)
 			}
 		}
@@ -146,9 +146,9 @@ func (cm *Matrix) SumTFPN(class int) {
 }
 
 func (cm *Matrix) ScoreClass(class int) {
-	tp := cm.TFPN.FloatValRowCell(class, 0)
-	fp := cm.TFPN.FloatValRowCell(class, 1)
-	fn := cm.TFPN.FloatValRowCell(class, 2)
+	tp := cm.TFPN.FloatValueRowCell(class, 0)
+	fp := cm.TFPN.FloatValueRowCell(class, 1)
+	fn := cm.TFPN.FloatValueRowCell(class, 2)
 
 	precision := tp / (tp + fp)
 	cm.ClassScores.SetFloatRowCell(class, 0, precision)
@@ -165,9 +165,9 @@ func (cm *Matrix) ScoreMatrix() {
 
 	n := cm.N.Len()
 	for i := 0; i < n; i++ {
-		tp += cm.TFPN.FloatValRowCell(i, 0)
-		fp += cm.TFPN.FloatValRowCell(i, 1)
-		fn += cm.TFPN.FloatValRowCell(i, 2)
+		tp += cm.TFPN.FloatValueRowCell(i, 0)
+		fp += cm.TFPN.FloatValueRowCell(i, 1)
+		fn += cm.TFPN.FloatValueRowCell(i, 2)
 	}
 
 	// micro F1 - ignores class
@@ -178,7 +178,7 @@ func (cm *Matrix) ScoreMatrix() {
 	// some classes might not have any instances so check NaN
 	f1 = 0.0
 	for i := 0; i < n; i++ {
-		classf1 := cm.ClassScores.FloatValRowCell(i, 2)
+		classf1 := cm.ClassScores.FloatValueRowCell(i, 2)
 		if math.IsNaN(classf1) == false {
 			f1 += classf1
 		}
@@ -190,11 +190,11 @@ func (cm *Matrix) ScoreMatrix() {
 	f1 = 0.0
 	totalN := 0.0
 	for i := 0; i < n; i++ {
-		classf1 := cm.ClassScores.FloatValRowCell(i, 2) * cm.N.FloatVal1D(i)
+		classf1 := cm.ClassScores.FloatValueRowCell(i, 2) * cm.N.FloatValue1D(i)
 		if math.IsNaN(classf1) == false {
 			f1 += classf1
 		}
-		totalN += cm.N.FloatVal1D(i)
+		totalN += cm.N.FloatValue1D(i)
 	}
 	cm.MatrixScores.SetFloat1D(2, f1/totalN)
 }

decoder/linear.go

@@ -40,7 +40,7 @@ type Linear struct {
 	Inputs []float32
 
 	// for holding layer values
-	ValsTsrs map[string]*etensor.Float32 `view:"-"`
+	ValuesTsrs map[string]*etensor.Float32 `view:"-"`
 
 	// synaptic weights: outer loop is units, inner loop is inputs
 	Weights etensor.Float32
@@ -61,7 +61,7 @@ type Linear struct {
 // Layer is the subset of emer.Layer that is used by this code
 type Layer interface {
 	Name() string
-	UnitValsTensor(tsr etensor.Tensor, varNm string, di int) error
+	UnitValuesTensor(tsr etensor.Tensor, varNm string, di int) error
 	Shape() *etensor.Shape
 }
 
@@ -182,15 +182,15 @@ func (dec *Linear) SetTargets(targs []float32) error {
 	return nil
 }
 
-// ValsTsr gets value tensor of given name, creating if not yet made
-func (dec *Linear) ValsTsr(name string) *etensor.Float32 {
-	if dec.ValsTsrs == nil {
-		dec.ValsTsrs = make(map[string]*etensor.Float32)
+// ValuesTsr gets value tensor of given name, creating if not yet made
+func (dec *Linear) ValuesTsr(name string) *etensor.Float32 {
+	if dec.ValuesTsrs == nil {
+		dec.ValuesTsrs = make(map[string]*etensor.Float32)
 	}
-	tsr, ok := dec.ValsTsrs[name]
+	tsr, ok := dec.ValuesTsrs[name]
 	if !ok {
 		tsr = &etensor.Float32{}
-		dec.ValsTsrs[name] = tsr
+		dec.ValuesTsrs[name] = tsr
 	}
 	return tsr
 }
@@ -201,8 +201,8 @@ func (dec *Linear) ValsTsr(name string) *etensor.Float32 {
 func (dec *Linear) Input(varNm string, di int) {
 	off := 0
 	for _, ly := range dec.Layers {
-		tsr := dec.ValsTsr(ly.Name())
-		ly.UnitValsTensor(tsr, varNm, di)
+		tsr := dec.ValuesTsr(ly.Name())
+		ly.UnitValuesTensor(tsr, varNm, di)
 		if dec.PoolIndex >= 0 {
 			shape := ly.Shape()
 			y := dec.PoolIndex / shape.Dim(1)

decoder/linear_test.go

@@ -21,7 +21,7 @@ func (tl *TestLayer) Name() string {
 	return "TestLayer"
 }
 
-func (tl *TestLayer) UnitValsTensor(tsr etensor.Tensor, varNm string, di int) error {
+func (tl *TestLayer) UnitValuesTensor(tsr etensor.Tensor, varNm string, di int) error {
 	src, ok := tl.tensors[varNm]
 	if !ok {
 		return fmt.Errorf("bad key: %s", varNm)

decoder/softmax.go

@@ -50,7 +50,7 @@ type SoftMax struct {
 	Target int
 
 	// for holding layer values
-	ValsTsrs map[string]*etensor.Float32 `view:"-"`
+	ValuesTsrs map[string]*etensor.Float32 `view:"-"`
 
 	// synaptic weights: outer loop is units, inner loop is inputs
 	Weights etensor.Float32
@@ -124,15 +124,15 @@ func (sm *SoftMax) TrainMPI(targ int) {
 	sm.BackMPI()
 }
 
-// ValsTsr gets value tensor of given name, creating if not yet made
-func (sm *SoftMax) ValsTsr(name string) *etensor.Float32 {
-	if sm.ValsTsrs == nil {
-		sm.ValsTsrs = make(map[string]*etensor.Float32)
+// ValuesTsr gets value tensor of given name, creating if not yet made
+func (sm *SoftMax) ValuesTsr(name string) *etensor.Float32 {
+	if sm.ValuesTsrs == nil {
+		sm.ValuesTsrs = make(map[string]*etensor.Float32)
 	}
-	tsr, ok := sm.ValsTsrs[name]
+	tsr, ok := sm.ValuesTsrs[name]
 	if !ok {
 		tsr = &etensor.Float32{}
-		sm.ValsTsrs[name] = tsr
+		sm.ValuesTsrs[name] = tsr
 	}
 	return tsr
 }
@@ -143,8 +143,8 @@ func (sm *SoftMax) ValsTsr(name string) *etensor.Float32 {
 func (sm *SoftMax) Input(varNm string, di int) {
 	off := 0
 	for _, ly := range sm.Layers {
-		tsr := sm.ValsTsr(ly.Name())
-		ly.UnitValsTensor(tsr, varNm, di)
+		tsr := sm.ValuesTsr(ly.Name())
+		ly.UnitValuesTensor(tsr, varNm, di)
 		for j, v := range tsr.Values {
 			sm.Inputs[off+j] = v
 		}

egui/gui.go

@@ -41,7 +41,7 @@ type GUI struct {
 	Grids map[string]*etview.TensorGrid
 
 	// the view update for managing updates of netview
-	ViewUpdt *netview.ViewUpdt `view:"-"`
+	ViewUpdate *netview.ViewUpdate `view:"-"`
 
 	// net data for recording in nogui mode, if !nil
 	NetData *netview.NetData `view:"-"`
@@ -90,7 +90,7 @@ func (gui *GUI) Stopped() {
 	if gui.Body == nil {
 		return
 	}
-	if gui.ViewUpdt != nil {
+	if gui.ViewUpdate != nil {
 		gui.UpdateNetViewWhenStopped()
 	}
 	gui.GoUpdateWindow()

egui/netview.go

@@ -12,16 +12,16 @@ import (
 
 // UpdateNetView updates the gui visualization of the network.
 func (gui *GUI) UpdateNetView() {
-	if gui.ViewUpdt != nil {
-		gui.ViewUpdt.Update()
+	if gui.ViewUpdate != nil {
+		gui.ViewUpdate.Update()
 	}
 }
 
 // UpdateNetViewWhenStopped updates the gui visualization of the network.
 // when stopped either via stepping or user hitting stop button.
 func (gui *GUI) UpdateNetViewWhenStopped() {
-	if gui.ViewUpdt != nil {
-		gui.ViewUpdt.UpdateWhenStopped()
+	if gui.ViewUpdate != nil {
+		gui.ViewUpdate.UpdateWhenStopped()
 	}
 }
 

egui/plots.go

@@ -55,7 +55,7 @@ func ConfigPlotFromLog(title string, plt *eplot.Plot2D, lg *elog.Logs, key etime
 		if item.Color != "" {
 			cp.Color = grr.Log1(colors.FromString(item.Color, nil))
 		}
-		cp.TensorIdx = item.TensorIdx
+		cp.TensorIndex = item.TensorIndex
 		cp.ErrCol = item.ErrCol
 
 		plt.Params.Title = title + " " + time + " Plot"

elog/README.md

@@ -132,7 +132,7 @@ There are various additional analysis functions called here, for example this on
 func (ss *Sim) LogRunStats() {
     sk := etime.Scope(etime.Train, etime.Run)
     lt := ss.Logs.TableDetailsScope(sk)
-    ix, _ := lt.NamedIdxView("RunStats")
+    ix, _ := lt.NamedIndexView("RunStats")
 
     spl := split.GroupBy(ix, []string{"Params"})
     split.Desc(spl, "FirstZero")
@@ -308,7 +308,7 @@ Computing stats on the principal components of variance (PCA) across different i
 // PCAStats computes PCA statistics on recorded hidden activation patterns
 // from Analyze, Trial log data
 func (ss *Sim) PCAStats() {
-    ss.Stats.PCAStats(ss.Logs.IdxView(etime.Analyze, etime.Trial), "ActM", ss.Net.LayersByClass("Hidden"))
+    ss.Stats.PCAStats(ss.Logs.IndexView(etime.Analyze, etime.Trial), "ActM", ss.Net.LayersByClass("Hidden"))
     ss.Logs.ResetLog(etime.Analyze, etime.Trial)
 }
 ```
@@ -348,7 +348,7 @@ Here's how you record the data and log the resulting stats, using the `Analyze`
 
 ## Error by Input Category
 
-This item creates a tensor column that records the average error for each category of input stimulus (e.g., for images from object categories), using the `split.GroupBy` function for `etable`.  The `IdxView` function (see also `NamedIdxView`) automatically manages the `etable.IdxView` indexed view onto a log table, which is used for all aggregation and further analysis of data, so that you can efficiently analyze filtered subsets of the original data.
+This item creates a tensor column that records the average error for each category of input stimulus (e.g., for images from object categories), using the `split.GroupBy` function for `etable`.  The `IndexView` function (see also `NamedIndexView`) automatically manages the `etable.IndexView` indexed view onto a log table, which is used for all aggregation and further analysis of data, so that you can efficiently analyze filtered subsets of the original data.
 
 ```Go
     ss.Logs.AddItem(&elog.Item{
@@ -358,10 +358,10 @@ This item creates a tensor column that records the average error for each catego
         DimNames:  []string{"Cat"},
         Plot:      true,
         Range:     minmax.F64{Min: 0},
-        TensorIdx: -1, // plot all values
+        TensorIndex: -1, // plot all values
         Write: elog.WriteMap{
             etime.Scope(etime.Test, etime.Epoch): func(ctx *elog.Context) {
-                ix := ctx.Logs.IdxView(etime.Test, etime.Trial)
+                ix := ctx.Logs.IndexView(etime.Test, etime.Trial)
                 spl := split.GroupBy(ix, []string{"Cat"})
                 split.AggTry(spl, "Err", agg.AggMean)
                 cats := spl.AggsToTable(etable.ColNameOnly)