Merge pull request #212 from opensource9ja/fix/iloc-slicing

Fix iloc slicing bug
javascriptdata · May 30, 2021 · 7af31b0 · 7af31b0
2 parents ffc598e + 08d16c5
commit 7af31b0
Show file tree

Hide file tree

Showing 14 changed files with 282 additions and 124 deletions.
diff --git a/danfojs-browser/lib/bundle.js b/danfojs-browser/lib/bundle.js
diff --git a/danfojs-browser/lib/bundle.js.map b/danfojs-browser/lib/bundle.js.map
diff --git a/danfojs-browser/src/core/indexing.js b/danfojs-browser/src/core/indexing.js
@@ -17,7 +17,6 @@ export const indexLoc = (ndframe, kwargs) => {
     if (Array.isArray(kwargs["rows"])) {
 
       if (kwargs["rows"].length == 1 && typeof kwargs["rows"][0] == "string") {
-        //console.log("here", kwargs["rows"].length)
         if (kwargs["rows"][0].includes(":")) {
 
           let column_split = kwargs["rows"][0].split(":");
@@ -93,16 +92,24 @@ export const indexLoc = (ndframe, kwargs) => {
         let column_split = kwargs["columns"][0].split(":");
         let start, end;
 
-        if (kwargs["type"] == "iloc" || (column_split[0] == "")) {
-          start = parseInt(column_split[0]) || 0;
-          end = parseInt(column_split[1]) - 1 === 0 ? 0 : parseInt(column_split[1]) - 1;
-        } else {
+        if (kwargs["type"] == "iloc") {
+          if (column_split[0] == "") {
+            start = 0;
+          } else {
+            start = parseInt(column_split[0]) || 0;
+          }
+
+          if (column_split[1] == "") {
+            end = parseInt(ndframe.columns.length - 1);
+          } else {
+            end = parseInt(column_split[1]) - 1 === 0 ? 0 : parseInt(column_split[1]) - 1;
+          }
 
+        } else {
           start = parseInt(ndframe.columns.indexOf(column_split[0]));
           end = parseInt(ndframe.columns.indexOf(column_split[1])) - 1;
         }
 
-
         if (typeof start == "number" && typeof end == "number") {
 
           columns = utils.__range(start, end);
@@ -136,7 +143,7 @@ export const indexLoc = (ndframe, kwargs) => {
       throw new Error(`Specified row index ${row_val} is bigger than maximum row index of ${max_rowIndex}`);
     }
 
-    if (Array.isArray(data_values[0])){
+    if (Array.isArray(data_values[0])) {
 
       let value = data_values[row_val];
       let row_data = [];

diff --git a/danfojs-browser/src/index.js b/danfojs-browser/src/index.js
@@ -1,9 +1,10 @@
 import NDframe from "./core/generic";
+import * as tf from "@tensorflow/tfjs";
 
 export { Series } from "./core/series";
 export { DataFrame } from "./core/frame";
 export { to_datetime } from "./core/timeseries";
-export { read_csv, read_json, read_excel } from "./io/reader";
+export { read_csv, read_json, read_excel, read } from "./io/reader";
 export { merge } from "./core/merge";
 export { concat } from "./core/concat";
 export { LabelEncoder, OneHotEncoder } from "./preprocessing/encodings";
@@ -14,6 +15,6 @@ export { Configs } from "./config/config";
 export { NDframe };
 export { Str } from "./core/strings";
 export { Utils } from "./core/utils";
-export * as tf from "@tensorflow/tfjs";
+export { tf };
 
-export const _version = "0.2.5";
+export const _version = "0.2.6";
diff --git a/danfojs-browser/tests/core/frame.js b/danfojs-browser/tests/core/frame.js
@@ -423,6 +423,68 @@ describe("DataFrame", function () {
 
     });
 
+    it("column slice starting with 0 and returning a single result works", function () {
+      let data = {
+        "Name": [ "Apples", "Mango", "Banana", "Pear" ],
+        "Count": [ 21, 5, 30, 10 ],
+        "Price": [ 200, 300, 40, 250 ]
+      };
+      let df = new dfd.DataFrame(data);
+      let sub_df = df.iloc({ rows: [ "2:3" ], columns: [ "0:1" ] });
+      const result = [ [ "Banana" ] ];
+      assert.deepEqual(sub_df.values, result);
+
+    });
+    it("column slice with format '0:' works", function () {
+      let data = {
+        "Name": [ "Apples", "Mango", "Banana", "Pear" ],
+        "Count": [ 21, 5, 30, 10 ],
+        "Price": [ 200, 300, 40, 250 ]
+      };
+      let df = new dfd.DataFrame(data);
+      let sub_df = df.iloc({ rows: [ "2:3" ], columns: [ "0:" ] });
+      const result = [ [ "Banana", 30, 40 ] ];
+      assert.deepEqual(sub_df.values, result);
+
+    });
+    it("column slice with format ':2' works", function () {
+      let data = {
+        "Name": [ "Apples", "Mango", "Banana", "Pear" ],
+        "Count": [ 21, 5, 30, 10 ],
+        "Price": [ 200, 300, 40, 250 ]
+      };
+      let df = new dfd.DataFrame(data);
+      let sub_df = df.iloc({ rows: [ "2:3" ], columns: [ ":2" ] });
+      const result = [ [ "Banana", 30 ] ];
+      assert.deepEqual(sub_df.values, result);
+
+    });
+    it("row slice with format ':2' works", function () {
+      let data = {
+        "Name": [ "Apples", "Mango", "Banana", "Pear" ],
+        "Count": [ 21, 5, 30, 10 ],
+        "Price": [ 200, 300, 40, 250 ]
+      };
+      let df = new dfd.DataFrame(data);
+      let sub_df = df.iloc({ rows: [ ":2" ], columns: [ ":1" ] });
+      const result = [ [ 'Apples' ], [ 'Mango' ] ];
+      assert.deepEqual(sub_df.values, result);
+
+    });
+    it("row slice with format '1:' works", function () {
+      let data = {
+        "Name": [ "Apples", "Mango", "Banana", "Pear" ],
+        "Count": [ 21, 5, 30, 10 ],
+        "Price": [ 200, 300, 40, 250 ]
+      };
+      let df = new dfd.DataFrame(data);
+      let sub_df = df.iloc({ rows: [ ":2" ], columns: [ ":2" ] });
+      const result = [ [ 'Apples', 21 ], [ 'Mango', 5 ] ];
+      assert.deepEqual(sub_df.values, result);
+
+    });
+
+
   });
 
 

diff --git a/danfojs-browser/tests/core/series.js b/danfojs-browser/tests/core/series.js
@@ -1223,6 +1223,18 @@ describe("Series", function () {
 
       assert.deepEqual(sf.iloc([ "1:4" ]).values, expected_val);
     });
+    it("indexing by slicing format ':5' works", function () {
+      let data = [ 1, 2, 3, 4, "a", "b", "c" ];
+      let sf = new dfd.Series(data);
+      let expected_val = [ 1, 2, 3, 4, 'a' ];
+      assert.deepEqual(sf.iloc([ ":5" ]).values, expected_val);
+    });
+    it("indexing by slicing format '2:` works", function () {
+      let data = [ 1, 2, 3, 4, "a", "b", "c" ];
+      let sf = new dfd.Series(data);
+      let expected_val = [ 3, 4, "a", "b", "c" ];
+      assert.deepEqual(sf.iloc([ "2:" ]).values, expected_val);
+    });
   });
 
   describe("append", function () {

diff --git a/danfojs-node/dist/core/indexing.js b/danfojs-node/dist/core/indexing.js
@@ -87,10 +87,18 @@ const indexLoc = (ndframe, kwargs) => {
         let column_split = kwargs["columns"][0].split(":");
         let start, end;
 
-        if (kwargs["type"] == "iloc" || column_split[0] == "") {
-          start = parseInt(column_split[0]) || 0;
-          end = parseInt(column_split[1]) - 1 === 0 ? 0 : parseInt(column_split[1]) - 1;
-          console.log(start, end);
+        if (kwargs["type"] == "iloc") {
+          if (column_split[0] == "") {
+            start = 0;
+          } else {
+            start = parseInt(column_split[0]) || 0;
+          }
+
+          if (column_split[1] == "") {
+            end = parseInt(ndframe.columns.length - 1);
+          } else {
+            end = parseInt(column_split[1]) - 1 === 0 ? 0 : parseInt(column_split[1]) - 1;
+          }
         } else {
           start = parseInt(ndframe.columns.indexOf(column_split[0]));
           end = parseInt(ndframe.columns.indexOf(column_split[1])) - 1;

diff --git a/danfojs-node/dist/index.js b/danfojs-node/dist/index.js
@@ -125,6 +125,10 @@ exports.tf = exports._version = void 0;
 
 var _generic = _interopRequireDefault(require("./core/generic"));
 
+var tf = _interopRequireWildcard(require("@tensorflow/tfjs-node"));
+
+exports.tf = tf;
+
 var _series = require("./core/series");
 
 var _frame = require("./core/frame");
@@ -151,8 +155,5 @@ var _strings = require("./core/strings");
 
 var _utils = require("./core/utils");
 
-var _tf = _interopRequireWildcard(require("@tensorflow/tfjs-node"));
-
-exports.tf = _tf;
 const _version = "0.2.6";
 exports._version = _version;
diff --git a/danfojs-node/dist/io/reader.js b/danfojs-node/dist/io/reader.js
@@ -79,7 +79,7 @@ const read = async (source, configs = {}) => {
   let rows, file;
 
   if ((0, _frictionless.isDataset)(source)) {
-    console.log("datapackage.json found. Loading Dataset package from Datahub.io");
+    console.info("datapackage.json found. Loading Dataset package from Datahub.io");
     const dataset = await _frictionless.Dataset.load(source);
     file = dataset.resources[data_num];
     rows = await (0, _streamToArray.default)(await file.rows());

diff --git a/danfojs-node/src/core/indexing.js b/danfojs-node/src/core/indexing.js
@@ -17,7 +17,6 @@ export const indexLoc = (ndframe, kwargs) => {
     if (Array.isArray(kwargs["rows"])) {
 
       if (kwargs["rows"].length == 1 && typeof kwargs["rows"][0] == "string") {
-        //console.log("here", kwargs["rows"].length)
         if (kwargs["rows"][0].includes(":")) {
 
           let column_split = kwargs["rows"][0].split(":");
@@ -93,17 +92,24 @@ export const indexLoc = (ndframe, kwargs) => {
         let column_split = kwargs["columns"][0].split(":");
         let start, end;
 
-        if (kwargs["type"] == "iloc" || (column_split[0] == "")) {
-          start = parseInt(column_split[0]) || 0;
-          end = parseInt(column_split[1]) - 1 === 0 ? 0 : parseInt(column_split[1]) - 1;
-          console.log(start, end);
-        } else {
+        if (kwargs["type"] == "iloc") {
+          if (column_split[0] == "") {
+            start = 0;
+          } else {
+            start = parseInt(column_split[0]) || 0;
+          }
+
+          if (column_split[1] == "") {
+            end = parseInt(ndframe.columns.length - 1);
+          } else {
+            end = parseInt(column_split[1]) - 1 === 0 ? 0 : parseInt(column_split[1]) - 1;
+          }
 
+        } else {
           start = parseInt(ndframe.columns.indexOf(column_split[0]));
           end = parseInt(ndframe.columns.indexOf(column_split[1])) - 1;
         }
 
-
         if (typeof start == "number" && typeof end == "number") {
 
           columns = utils.__range(start, end);
@@ -137,7 +143,7 @@ export const indexLoc = (ndframe, kwargs) => {
       throw new Error(`Specified row index ${row_val} is bigger than maximum row index of ${max_rowIndex}`);
     }
 
-    if (Array.isArray(data_values[0])){
+    if (Array.isArray(data_values[0])) {
 
       let value = data_values[row_val];
       let row_data = [];
@@ -188,5 +194,5 @@ export const indexLoc = (ndframe, kwargs) => {
   rows.forEach((i) => {
     final_row.push(ndframe.index[i]);
   });
-  return [ new_data, column_names, final_row ];
+  return [new_data, column_names, final_row];
 };
diff --git a/danfojs-node/src/index.js b/danfojs-node/src/index.js
@@ -1,4 +1,5 @@
 import NDframe from "./core/generic";
+import * as tf from "@tensorflow/tfjs-node";
 
 export { Series } from "./core/series";
 export { DataFrame } from "./core/frame";
@@ -14,6 +15,6 @@ export { Configs } from "./config/config";
 export { NDframe };
 export { Str } from "./core/strings";
 export { Utils } from "./core/utils";
-export * as tf from "@tensorflow/tfjs-node";
+export { tf };
 
 export const _version = "0.2.6";
diff --git a/danfojs-node/src/io/reader.js b/danfojs-node/src/io/reader.js
@@ -103,7 +103,7 @@ export const read = async (
   let rows, file;
 
   if (isDataset(source)) {
-    console.log(
+    console.info(
       "datapackage.json found. Loading Dataset package from Datahub.io"
     );
     const dataset = await Dataset.load(source);