diff --git a/examples/Dataset_Profiler_Viewer.ipynb b/examples/Dataset_Profiler_Viewer.ipynb
index 48b2c8809b..768ba8de08 100644
--- a/examples/Dataset_Profiler_Viewer.ipynb
+++ b/examples/Dataset_Profiler_Viewer.ipynb
@@ -18,7 +18,7 @@
},
{
"cell_type": "code",
- "execution_count": null,
+ "execution_count": 1,
"id": "1f0ab34a",
"metadata": {},
"outputs": [],
@@ -29,7 +29,7 @@
},
{
"cell_type": "code",
- "execution_count": null,
+ "execution_count": 2,
"id": "fc04cc42",
"metadata": {},
"outputs": [],
@@ -59,7 +59,7 @@
},
{
"cell_type": "code",
- "execution_count": 14,
+ "execution_count": 4,
"id": "7ebc0b5f",
"metadata": {},
"outputs": [],
@@ -86,22 +86,24 @@
},
{
"cell_type": "code",
- "execution_count": 17,
+ "execution_count": null,
"id": "2e018520",
"metadata": {},
"outputs": [],
"source": [
"session = get_or_create_session()\n",
- "with session.logger(\"mytestytest\",dataset_timestamp=datetime.datetime(2021, 6, 2)) as logger:\n",
- " for _ in range(500):\n",
- " logger.log({\"uniform_integers\": np.random.randint(0,50)})\n",
- " logger.log({\"strings\": fake.name()})\n",
- " logger.log({\"mixture_distribution\": np.random.choice(distribution, 1)[0]})\n",
- " logger.log({\"nulls\": None})\n",
- " logger.log({\"moah_data\":1})\n",
- " logger.log({\"moah_data\":1})\n",
- " logger.log({\"moah_data\":5})\n",
- " profile=logger.profile"
+ "def profile_generator():\n",
+ " with session.logger(\"mytestytest\",dataset_timestamp=datetime.datetime(2021, 6, 2)) as logger:\n",
+ " for _ in range(500):\n",
+ " logger.log({\"uniform_integers\": np.random.randint(0,50)})\n",
+ " logger.log({\"strings\": fake.name()})\n",
+ " logger.log({\"mixture_distribution\": np.random.choice(distribution, 1)[0]})\n",
+ " logger.log({\"nulls\": None})\n",
+ " logger.log({\"moah_data\":1})\n",
+ " logger.log({\"moah_data\":1})\n",
+ " logger.log({\"moah_data\":5})\n",
+ " return logger.profile\n",
+ "profile=profile_generator()"
]
},
{
@@ -123,29 +125,54 @@
},
{
"cell_type": "code",
- "execution_count": 16,
+ "execution_count": null,
"id": "e293634d",
"metadata": {},
- "outputs": [
- {
- "data": {
- "text/plain": [
- "'/var/folders/pr/f715zv8x17b1v5vwydgv2gq40000gq/T/tmpabxku502.html'"
- ]
- },
- "execution_count": 16,
- "metadata": {},
- "output_type": "execute_result"
- }
- ],
+ "outputs": [],
"source": [
"profile_viewer(profiles=[profile], output_path=None)"
]
},
+ {
+ "cell_type": "markdown",
+ "id": "e9cebe5c",
+ "metadata": {},
+ "source": [
+ "### Logging reference profile"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": 8,
+ "id": "d36563e9",
+ "metadata": {},
+ "outputs": [],
+ "source": [
+ "reference_profile=profile_generator() "
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "id": "56dcaef9",
+ "metadata": {},
+ "source": [
+ "### Add reference profile to viewer and open"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": null,
+ "id": "6a4eb38c",
+ "metadata": {},
+ "outputs": [],
+ "source": [
+ "profile_viewer(profiles=[profile], reference_profiles=[reference_profile], output_path=None,)"
+ ]
+ },
{
"cell_type": "code",
"execution_count": null,
- "id": "d09732c5",
+ "id": "1b1d89e2",
"metadata": {},
"outputs": [],
"source": []
@@ -153,9 +180,9 @@
],
"metadata": {
"kernelspec": {
- "display_name": "whylogs-dev",
+ "display_name": "Python 3",
"language": "python",
- "name": "whylogs-dev"
+ "name": "python3"
},
"language_info": {
"codemirror_mode": {
@@ -167,7 +194,7 @@
"name": "python",
"nbconvert_exporter": "python",
"pygments_lexer": "ipython3",
- "version": "3.9.6"
+ "version": "3.8.8"
}
},
"nbformat": 4,
diff --git a/examples/Profile_Viewer_In_Notebook.ipynb b/examples/Profile_Viewer_In_Notebook.ipynb
new file mode 100644
index 0000000000..ee2876a96a
--- /dev/null
+++ b/examples/Profile_Viewer_In_Notebook.ipynb
@@ -0,0 +1,503 @@
+{
+ "cells": [
+ {
+ "cell_type": "markdown",
+ "id": "6b36ea8e",
+ "metadata": {},
+ "source": [
+ "# 📈 Whylogs Profile Visualization\n",
+ "### gives you various ways to simplify the proccess of detecting dataset drift. Bellow you can see the instructions for creating dummy dataset and list of currently available visual reports."
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "id": "be56fca7",
+ "metadata": {},
+ "source": [
+ "### 🗂️Install dependencies and make imports"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": null,
+ "id": "29e48da5",
+ "metadata": {},
+ "outputs": [],
+ "source": [
+ "!pip install faker\n",
+ "!pip install pybars3"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": null,
+ "id": "5d40e474",
+ "metadata": {},
+ "outputs": [],
+ "source": [
+ "import numpy as np\n",
+ "import pandas as pd\n",
+ "import datetime\n",
+ "from collections import OrderedDict\n",
+ "from faker import Faker"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": null,
+ "id": "2c3ff3a2",
+ "metadata": {},
+ "outputs": [],
+ "source": [
+ "from whylogs import get_or_create_session\n",
+ "from whylogs.viz import NotebookProfileViewer\n",
+ "from whylogs.core.statistics.constraints import (\n",
+ " columnValuesInSetConstraint,\n",
+ " containsEmailConstraint,\n",
+ " minBetweenConstraint,\n",
+ " maxLessThanEqualConstraint,\n",
+ " parametrizedKSTestPValueGreaterThanConstraint,\n",
+ " columnsMatchSetConstraint,\n",
+ " columnPairValuesInSetConstraint,\n",
+ " sumOfRowValuesOfMultipleColumnsEqualsConstraint,\n",
+ " columnValuesUniqueWithinRow,\n",
+ " DatasetConstraints\n",
+ ")"
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "id": "272f6054",
+ "metadata": {},
+ "source": [
+ "### ♻️Create dummy data"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": null,
+ "id": "41f8a0f9",
+ "metadata": {},
+ "outputs": [],
+ "source": [
+ "locales = OrderedDict([\n",
+ " ('en-US', 1),\n",
+ " ('fr-FR', 2),\n",
+ " ('ja_JP', 2),\n",
+ "])\n",
+ "fake = Faker(locales)\n",
+ "distribution = np.concatenate((np.random.normal(0.1, 0.1, 500), np.random.normal(0.6, 0.2, 500)))"
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "id": "200841f2",
+ "metadata": {},
+ "source": [
+ "### 📝 Log it\n"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": null,
+ "id": "ea4ce6a6",
+ "metadata": {},
+ "outputs": [],
+ "source": [
+ "session = get_or_create_session()\n",
+ "def profile_generator():\n",
+ " with session.logger(\"mytestytest\", dataset_timestamp=datetime.datetime(2021, 6, 2)) as logger:\n",
+ " for _ in range(500):\n",
+ " logger.log({\"uniform_integers\": np.random.randint(0,50)})\n",
+ " logger.log({\"strings\": fake.name()})\n",
+ " logger.log({\"mixture_distribution\": np.random.choice(distribution, 1)[0]}) \n",
+ " logger.log({\"1mixture_distribution\": np.random.choice(distribution, 1)[0]})\n",
+ " logger.log({\"2mixture_distribution\": np.random.choice(distribution, 1)[0]})\n",
+ " logger.log({\"3mixture_distribution\": np.random.choice(distribution, 1)[0]})\n",
+ " logger.log({\"4mixture_distribution\": np.random.choice(distribution, 1)[0]})\n",
+ " logger.log({\"nulls\": None})\n",
+ " logger.log({\"moah_data\": 1})\n",
+ " logger.log({\"moah_data\": 1})\n",
+ " logger.log({\"moah_data\": 5})\n",
+ "\n",
+ " return logger.profile\n",
+ " \n",
+ "target_profile = profile_generator()\n",
+ "\n",
+ "reference_profile = profile_generator()"
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "id": "8390f360",
+ "metadata": {},
+ "source": [
+ "## ✨ Vizualize profiles with Whylogs"
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "id": "562fbacf",
+ "metadata": {},
+ "source": [
+ "### Initialization\n",
+ "Initialize Profile viewer by passing profiles for which you want to get the visualizations"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": null,
+ "id": "7dbff14f",
+ "metadata": {},
+ "outputs": [],
+ "source": [
+ "visualization = NotebookProfileViewer()\n",
+ "visualization.set_profiles(target_profile=target_profile, reference_profile=reference_profile)"
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "id": "87ea53b6",
+ "metadata": {},
+ "source": [
+ "###### `*target_profiles`: Profiled dataset which will be reffered as `target`\n",
+ "###### `*reference_profiles`: Profiled dataset which will be reffered as `reference`"
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "id": "9908a9f9",
+ "metadata": {},
+ "source": [
+ "### Summary Drift Report"
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "id": "650d44c4",
+ "metadata": {},
+ "source": [
+ "You can get summary drift report for `target` and `reference` profiles features"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": null,
+ "id": "911147eb",
+ "metadata": {
+ "scrolled": true
+ },
+ "outputs": [],
+ "source": [
+ "visualization.summary_drift_report(preferred_cell_height=\"1000px\")"
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "id": "c4be405d",
+ "metadata": {},
+ "source": [
+ "###### `preferred_cell_height`: height in `px` for generated visualization cell "
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "id": "4dd9e690",
+ "metadata": {},
+ "source": [
+ "### Double histogram"
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "id": "28b512ca",
+ "metadata": {},
+ "source": [
+ "You can get double histogram for numerical features"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": null,
+ "id": "250bb662",
+ "metadata": {
+ "scrolled": true
+ },
+ "outputs": [],
+ "source": [
+ "visualization.double_histogram(feature_names=\"uniform_integers\")"
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "id": "af84a332",
+ "metadata": {},
+ "source": [
+ "###### `*feature_names`: string or list of strings containing names of the features for which you want to see double histogram\n",
+ "###### `preferred_cell_height`: height in `px` for generated visualization cell "
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "id": "193e306d",
+ "metadata": {},
+ "source": [
+ "### Distribution chart"
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "id": "6f4931fb",
+ "metadata": {},
+ "source": [
+ "You can get distirubtion chart for categorical features"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": null,
+ "id": "ccc7c73e",
+ "metadata": {
+ "scrolled": true
+ },
+ "outputs": [],
+ "source": [
+ "visualization.distribution_chart(feature_names=\"strings\")"
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "id": "cd97db45",
+ "metadata": {},
+ "source": [
+ "###### `*feature_names`: string or list of strings containing names of the features for which you want to see double histogram\n",
+ "###### `preferred_cell_height`: height in `px` for generated visualization cell "
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "id": "408e1c5b",
+ "metadata": {},
+ "source": [
+ "### Differenced distribution chart"
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "id": "4d65b0b4",
+ "metadata": {},
+ "source": [
+ "You can get differenced distirubtion chart for categorical features"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": null,
+ "id": "d26b2fa2",
+ "metadata": {
+ "scrolled": true
+ },
+ "outputs": [],
+ "source": [
+ "visualization.difference_distribution_chart(feature_names=\"strings\")"
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "id": "92451fbf",
+ "metadata": {},
+ "source": [
+ "###### `*feature_names`: string or list of strings containing names of the features for which you want to see double histogram\n",
+ "###### `preferred_cell_height`: height in `px` for generated visualization cell "
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "id": "78db1538",
+ "metadata": {},
+ "source": [
+ "### Feature Statistics"
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "id": "4dc0e3f6",
+ "metadata": {},
+ "source": [
+ "You can get set of useful statistics for features by passing the profile and feature names"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": null,
+ "id": "fb77bcd5",
+ "metadata": {
+ "scrolled": false
+ },
+ "outputs": [],
+ "source": [
+ "visualization.feature_statistics(feature_name=\"mixture_distribution\", profile=\"reference\")"
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "id": "8b483a1d",
+ "metadata": {},
+ "source": [
+ "###### `*feature_name`: Any feature name from your profiled dataset\n",
+ "###### `profile_name`: `\"target\"` or `\"reference\"`\n",
+ "###### `prefered_cell_height`: height in `px` for generated visualization cell "
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "id": "a3469ba2",
+ "metadata": {},
+ "source": [
+ "### Generate constraints"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": null,
+ "id": "dafab41d",
+ "metadata": {},
+ "outputs": [],
+ "source": [
+ "def get_sample_dataset_constraints():\n",
+ " cvisc = columnValuesInSetConstraint(value_set={2, 5, 8})\n",
+ " email_constraint = containsEmailConstraint()\n",
+ "\n",
+ " min_gt_constraint = minBetweenConstraint(lower_value=1, upper_value=5)\n",
+ " max_le_constraint = maxLessThanEqualConstraint(value=100)\n",
+ "\n",
+ " distribution = np.random.normal(0, 1, 50)\n",
+ "\n",
+ " ks_test_p_value_constraint = parametrizedKSTestPValueGreaterThanConstraint(\n",
+ " distribution,\n",
+ " p_value=0.5,\n",
+ " name=\"has a standard normal distribution\"\n",
+ " )\n",
+ "\n",
+ " set1 = set([\"col1\", \"col2\"])\n",
+ " columns_match_constraint = columnsMatchSetConstraint(set1)\n",
+ "\n",
+ " val_set = {(1, 2), (3, 5)}\n",
+ " col_set = [\"A\", \"B\"]\n",
+ " mcv_constraints = [\n",
+ " columnPairValuesInSetConstraint(column_A=\"A\", column_B=\"B\", value_set=val_set),\n",
+ " sumOfRowValuesOfMultipleColumnsEqualsConstraint(columns=col_set, value=100),\n",
+ " columnValuesUniqueWithinRow(column_A=\"A\", verbose=True),\n",
+ " ]\n",
+ "\n",
+ " return DatasetConstraints(\n",
+ " None,\n",
+ " value_constraints={\"A\": [cvisc], \"users\": [email_constraint]},\n",
+ " summary_constraints={\"B\": [max_le_constraint, min_gt_constraint], \"value\": [ks_test_p_value_constraint]},\n",
+ " table_shape_constraints=[columns_match_constraint],\n",
+ " multi_column_value_constraints=mcv_constraints,\n",
+ " )\n",
+ "\n",
+ "data = pd.DataFrame({\n",
+ " \"A\": [1, 2, 2, 5, 7, 6],\n",
+ " \"B\": [5, 4, 5, 1, 6, 0],\n",
+ " \"users\": [\"john\", \"jane@example.com\", \"alex\", \"bob\", \"anna@example.com\", \"dave\"],\n",
+ " \"value\": [23.4, 123.2, 423.3, 32.1, 42.2, 344.2],\n",
+ "})\n",
+ "\n",
+ "dc = get_sample_dataset_constraints()\n",
+ "constraints_profile = session.log_dataframe(data, \"test.data\", constraints=dc)\n",
+ "constraints_profile.apply_summary_constraints()\n",
+ "constraints_profile.apply_table_shape_constraints()\n",
+ "session.close()"
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "id": "17ac9327",
+ "metadata": {},
+ "source": [
+ "### Constraints report"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": null,
+ "id": "4a2786f8",
+ "metadata": {},
+ "outputs": [],
+ "source": [
+ "visualization.constraints_report(dc)"
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "id": "b04f735b",
+ "metadata": {},
+ "source": [
+ "### Download prefered cell output"
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "id": "2015ff5d",
+ "metadata": {},
+ "source": [
+ "You can also download any of those visualisation in `HTML` format for further analysys, by passing the visualization name"
+ ]
+ },
+ {
+ "cell_type": "code",
+ "execution_count": null,
+ "id": "980d7236",
+ "metadata": {},
+ "outputs": [],
+ "source": [
+ "visualization.download(html=visualization.summary_drift_report(), html_file_name='example')"
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "id": "01fbed2c",
+ "metadata": {},
+ "source": [
+ "By calling `download()` method of `DisplayProfile` and passing visualizer command, path to be downloaded to (optional) and name of the file you prefer (optional). Exaple `download(visualization.feature(\"title\"), path=\"examlpe/path\", html_file_name=\"example_html_file_name\")`. Command will download HTML format.\n",
+ "\n",
+ "If path is not passed file will be downloaded to `html_reports` located in whylogs directory by default.\n",
+ "\n",
+ "If name of the file is not passed it will be name of the dataset followed by timestamp of the profile by default. "
+ ]
+ },
+ {
+ "cell_type": "markdown",
+ "id": "7ffc293c",
+ "metadata": {},
+ "source": [
+ "###### `*feature_name`: Any feature name from your profiled dataset\n",
+ "###### `preferred_path`: save path `default:` `/html_reports` located in whylogs directory\n",
+ "###### `html_file_name`: name of the file `default:` name of the dataset followed by timestamp of the profile"
+ ]
+ }
+ ],
+ "metadata": {
+ "kernelspec": {
+ "display_name": "Python 3 (ipykernel)",
+ "language": "python",
+ "name": "python3"
+ },
+ "language_info": {
+ "codemirror_mode": {
+ "name": "ipython",
+ "version": 3
+ },
+ "file_extension": ".py",
+ "mimetype": "text/x-python",
+ "name": "python",
+ "nbconvert_exporter": "python",
+ "pygments_lexer": "ipython3",
+ "version": "3.8.10"
+ }
+ },
+ "nbformat": 4,
+ "nbformat_minor": 5
+}
diff --git a/html_reports/.gitignore b/html_reports/.gitignore
new file mode 100644
index 0000000000..2d19fc766d
--- /dev/null
+++ b/html_reports/.gitignore
@@ -0,0 +1 @@
+*.html
diff --git a/src/whylogs/viewer/css/whylogs-styles.css b/src/whylogs/viewer/css/whylogs-styles.css
index c7c41eea94..f5c06a4119 100644
--- a/src/whylogs/viewer/css/whylogs-styles.css
+++ b/src/whylogs/viewer/css/whylogs-styles.css
@@ -154,6 +154,7 @@
z-index: 1000;
display: block;
padding: 20px;
+ padding-bottom: 0;
overflow-x: hidden;
overflow-y: auto; /* Scrollable contents if viewport is shorter than content. */
background-color: var(--brandSecondary100);
@@ -196,7 +197,7 @@
margin-top: 0;
}
-/*
+/*
* Custom styles
*/
@@ -374,7 +375,6 @@ a .wl-table-row--bottom-shadow {
.wl-table-cell,
.wl-table-head {
- border-right: 1px solid var(--brandSecondary200);
border-bottom: 1px solid var(--brandSecondary200);
display: table-cell;
padding: 12px 18px;
@@ -452,9 +452,12 @@ a .wl-table-row--bottom-shadow {
white-space: nowrap;
overflow: hidden;
font-style: italic;
+ text-align: center;
color: var(--brandSecondary400);
}
.wl-table-cell__bedge {
+ display: flex;
+ align-items: center;
height: 24px;
margin: 1px;
padding: 2px 8px;
@@ -668,15 +671,6 @@ a .wl-table-row--bottom-shadow {
justify-content: center;
}
-@media screen and (min-width: 1000px) {
- .desktop-content {
- display: block;
- }
- .no-responsive {
- display: none;
- }
-}
-
.no-responsive__content {
max-width: 600px;
width: 100%;
@@ -697,3 +691,558 @@ a .wl-table-row--bottom-shadow {
color: var(--brandSecondary900);
line-height: 1.5;
}
+
+.wl-compare-profile {
+ position: sticky;
+ left: 0;
+ display: flex;
+ padding: 18px;
+ justify-content: flex-end;
+ flex-direction: row-reverse;
+}
+
+.wl-property-panel__chart--single {
+ padding-bottom: 17px;
+}
+
+.wl-property-panel__chart-title {
+ color: var(--brandSecondary900);
+ font-size: 14px;
+ font-family: Asap, sans-serif;
+ font-weight: bold;
+ margin-bottom: 15px;
+}
+
+.wl-select-reference-profile {
+ display: flex;
+ align-items: center;
+}
+
+.wl__unlock-the-power {
+ padding: 15px;
+ position: fixed;
+ bottom: 0;
+ left: 0;
+ z-index: 0;
+ background: #F1F6F6;
+ margin-bottom: 0 !important;
+ display: flex;
+ flex-direction: column;
+ align-items: center;
+ -webkit-box-shadow: 0px -6px 13px 0px rgba(0,0,0,0.32);
+ box-shadow: 0px 0px 15px rgba(0, 0, 0, 0.05);
+ width: var(--SIDE-PANEL-WIDTH);
+}
+
+.wl__btn-signup {
+ font-size: 13px;
+ padding: 13px;
+ padding-left: 20px;
+ padding-right: 20px;
+}
+
+.wl__close-icon {
+ width: 15px;
+ height: 15px;
+}
+
+.wl__close-icon img{
+ width: 100%;
+ height: 100%;
+ cursor: pointer;
+}
+
+.count-color {
+ color: black;
+}
+
+.wl__dropdown_arrow-icon_container {
+ width: 18px;
+ height: 16px;
+}
+
+.wl__dropdown_arrow-icon_container img{
+ width: 100%;
+ height: 100%;
+ cursor: pointer;
+}
+
+.wl__dropdown_arrow-icon, .open-sign-up-text{
+ position: relative;
+}
+
+.notif-circle-container, .open-sign-up-text-notif-container{
+ position: absolute;
+ top: -5px;
+ right: -5px;
+ padding: 5.3px;
+ border-radius: 50%;
+ background-color: var(--brandSecondary100);
+ cursor: pointer;
+}
+
+.open-sign-up-text-notif-container {
+ padding: 10.3px;
+}
+
+.notif-circle {
+ position: absolute;
+ top: 2px;
+ right: 2px;
+ padding: 3.3px;
+ border-radius: 50%;
+ background-color: #F2994A;
+}
+
+.open-sign-up-text-notif {
+ display: flex;
+ position: absolute;
+ top: 2px;
+ right: 2px;
+ padding: 2.5px 6px;
+ color: #FFF;
+ border-radius: 50%;
+ background-color: #5CAEBC;
+}
+
+.space-between {
+ display: flex;
+ justify-content: space-between;
+}
+
+.feature-count-title {
+ font-size: 22px;
+}
+
+.wl_filter-options {
+ background: rgba(255, 255, 255, 0.5);
+ border: 1px solid #DBE5E7;
+ box-sizing: border-box;
+ border-radius: 4px;
+ padding: 10px;
+}
+
+.form-check-input:checked {
+ background-color: #0E7384;
+ border-color: #0E7384;
+}
+
+.form-check-input[type=checkbox] {
+ border-radius: 1.25em;
+}
+
+.not-clickable {
+ pointer-events: none;
+}
+
+.wl_arrow-icon {
+ width: 15px;
+ height: 15px;
+}
+
+.arrow-icon-container {
+ height: 100%;
+ cursor: pointer;
+}
+
+.wl_list-item-dot {
+ background: #b0d2d7;
+ width: 8px;
+ height: 8px;
+ border-radius: 50px;
+}
+
+.wl_filter-list-item {
+ display: flex;
+ align-items: center;
+}
+
+.wl_filter-list-item>span{
+ padding-left: 15px;
+}
+
+.display-flex{
+ display: flex;
+}
+
+.table-border-none {
+ padding: 0;
+ border: none;
+}
+
+.clickable-test-feature-wrap {
+ background: #F8FAFB;
+}
+
+.clickable-test-feature-heading {
+ display: flex;
+ flex-direction: column;
+}
+
+.clickable-test-feature-heading-wrap {
+ padding: 35px;
+ padding-bottom: 0;
+ border-bottom: 2px solid #EBF2F3;
+}
+
+.pages-button-wrap {
+ display: flex;
+ align-items: flex-end;
+}
+
+.page-button-wrap {
+ padding-right: 60px;
+}
+
+.page-button{
+ border: none;
+ background: no-repeat;
+ color: #6C757D;
+ font-weight: 600;
+ font-size: 16px;
+ letter-spacing: -0.01em;
+ padding-bottom: 25px;
+}
+
+.activ-pages-button{
+ color: #369BAC !important;
+ border-bottom: 4px solid #369BAC !important;
+ padding-bottom: 21px !important;
+}
+
+.title p {
+ font-weight: 600;
+ font-size: 24px;
+ color: #313B3D;
+}
+
+.info p {
+ font-weight: 600;
+ font-size: 12px;
+ color: #313B3D;
+}
+
+.info div {
+ font-weight: 600;
+ font-size: 18px;
+ line-height: 20px;
+ color: #0E7384;
+}
+
+.chart, .info {
+ display: flex;
+ flex-direction: column;
+ align-items: flex-end;
+ padding-right: 20px;
+}
+
+.chart {
+ margin-bottom: 1rem;
+}
+
+.info:last-child {
+ padding: 0;
+}
+
+.clickable-test-feature-body {
+ display: flex;
+ flex-direction: column;
+}
+
+.chart-box-wrap {
+ display: flex;
+ justify-content: center;
+}
+
+.chart-box {
+ display: flex;
+ align-items: center;
+ justify-content: center;
+ flex-direction: column;
+ width: 75%;
+ height: 310px;
+ border: 2px solid #EBF2F3;
+ background: #FFF;
+ border-radius: 4px;
+}
+
+.chart-box-title {
+ width: 80%;
+ justify-content: space-between;
+}
+
+.chart-box-title p{
+ font-family: Asap;
+ font-weight: bold;
+ font-size: 18px;
+ line-height: 16px;
+ color: #4F595B;
+}
+
+.chart-info-wrap {
+ display: flex;
+ align-items: flex-end;
+}
+
+.chart-info {
+ display: flex;
+ align-items: flex-end;
+}
+
+.property-panel-close-icon {
+ cursor: pointer;
+}
+
+.reference-json-form {
+ padding-right: 30px;
+}
+
+.flex-direction-colum {
+ display: flex;
+ flex-direction: column;
+}
+
+.feature-file-name {
+ font-size: 17px;
+ font-weight: 900;
+ color: #6C757D;
+}
+
+.dropdown-container {
+ background: rgba(255, 255, 255, 0.5) !important;
+ padding: 10px !important;
+ border: none !important;
+}
+
+.dropdown {
+ align-items: center;
+}
+
+.align-items {
+ align-items: center;
+}
+
+.dropdown p {
+ font-family: Asap;
+ font-size: 15px;
+ line-height: 150%;
+ margin: 0;
+}
+.dropdown img {
+ height: 10px;
+ cursor: pointer;
+}
+
+.search-input{
+ padding-top: 0 !important;
+ padding-bottom: 0 !important;
+}
+
+.search-input input{
+ border: none;
+ background: none;
+ outline: none;
+ height: 40px;
+ width: 100%;
+}
+
+.search-input img{
+ height: 19px;
+ pointer-events: none;
+}
+
+.edit-button-wrap {
+ padding-right: 5px;
+ cursor: pointer;
+}
+
+.edit-button {
+ border: 1px solid #369BAC;
+ box-sizing: border-box;
+ border-radius: 3px;
+ width: 33px;
+ text-align: center;
+}
+
+.edit-button label{
+ color: #369BAC;
+ font-family: Asap;
+ font-weight: 500;
+ line-height: 11px;
+ cursor: pointer;
+}
+
+.remove-reference-profile-button {
+ display: flex;
+ align-items: center;
+ justify-content: center;
+ border: 1px solid #6C757D;
+ border-radius: 3px;
+ cursor: pointer;
+}
+
+.remove-reference-profile-button img {
+ padding: 3px
+}
+
+input::placeholder {
+ color: var(--secondaryLight1000);
+}
+
+.wl-selected-profile {
+ position: sticky;
+ left: 0;
+}
+
+.circle-color {
+ display: inline-block;
+ padding: 5px;
+ border-radius: 50px;
+}
+
+.colors-for-distingushing-charts {
+ padding-right: 10px;
+}
+
+.summary-statistic {
+ padding-right: 20px;
+}
+
+.summary {
+ font-family: Asap;
+ font-style: normal;
+ font-weight: 600;
+ font-size: 14px;
+ line-height: 16px;
+ color: #6C757D;
+ padding-right: 10px;
+}
+
+.summary-title {
+ font-family: Asap;
+ font-weight: normal;
+ font-size: 15px;
+ line-height: 14px;
+ color: #6C757D;
+}
+
+.wl-selected-reference-profile {
+ font-weight: 900;
+}
+
+.reference-profile-time-stamp{
+ padding-left: 20px;
+}
+
+.reference-profile-time-stamp>strong{
+ font-size: 1rem;
+}
+
+.bar.positive {
+ fill: #369BACB2;
+}
+
+.bar.negative {
+ fill: #2683C9E5;
+}
+
+.frequent-item-box {
+ padding: 20px;
+ width: 90%;
+ align-items: flex-start;
+ display: flex;
+ justify-content: center;
+ flex-direction: column;
+ border: 2px solid #EBF2F3;
+ background: #FFF;
+ border-radius: 4px;
+}
+
+.frequent-item-box-wrap {
+ width: 100%;
+ display: flex;
+ justify-content: center;
+}
+
+.frequent-item-box-to-title {
+ margin-left: 30px;
+ margin-bottom: 20px;
+ font-family: Asap;
+ font-weight: bold;
+ font-size: 16px;
+ line-height: 16px;
+ color: #4F595B;
+}
+
+.frequent-items-body {
+ display: flex;
+ flex-direction: column;
+}
+
+.text-align-center {
+ text-align: center;
+}
+
+.fequent-items-wrap {
+ width: 100%;
+}
+
+.svg-container {
+ display: inline-block;
+ position: relative;
+ width: 85%;
+ padding-bottom: 34%;
+ vertical-align: top;
+ overflow: hidden;
+}
+.svg-content-responsive {
+ display: inline-block;
+ position: absolute;
+ left: 0;
+}
+
+.graph-svg-container {
+ padding-bottom: 12%;
+}
+
+.reference-table-head {
+ min-width: 305px;
+}
+
+@media screen and (min-width: 1000px) {
+ .desktop-content {
+ display: block;
+ }
+ .no-responsive {
+ display: none;
+ }
+ .wl__unlock-the-power {
+ z-index: 999999;
+ }
+}
+
+@media only screen and (min-width: 1350px) {
+ .clickable-test-feature-body {
+ display: flex;
+ flex-direction: row;
+ flex-wrap: wrap;
+ justify-content: center;
+ row-gap: 20px;
+ column-gap: 20px;
+ }
+
+ .chart-box-wrap {
+ width: 45%;
+ display: flex;
+ justify-content: center;
+ margin-bottom: 0 !important;
+ }
+
+ .chart-box {
+ width: 100%;
+ height: 310px;
+ border: 2px solid #EBF2F3;
+ background: #FFF;
+ border-radius: 4px;
+ }
+}
diff --git a/src/whylogs/viewer/index-hbs-cdn-all-in.html b/src/whylogs/viewer/index-hbs-cdn-all-in.html
deleted file mode 100644
index 40141b4c37..0000000000
--- a/src/whylogs/viewer/index-hbs-cdn-all-in.html
+++ /dev/null
@@ -1,1407 +0,0 @@
-
-
-
- Profile Viewer | whylogs
-
- Profile Viewer | whylogs
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
-
Feature
-
Frequent items
-
Inf. feature type
-
Total count
-
Null fraction
-
Est. unique values
-
Data type
-
Data type count
-
Mean
-
Std. dev
-
Min
-
First quantile
-
Median
-
Third quantile
-
Max
-
-
-
-
-
-
-
-
Please select and load a JSON profile from your local drive.
-
-
-
-
-
-
-
-
-
- Item
-
-
-
-
-
-
-
-
-
-
-
-
-
Hold on! :)
-
- It looks like your current screen size or device is not yet supported by the WhyLabs Sandbox. The Sandbox is
- best experienced on a desktop computer. Please try maximizing this window or switching to another device. We
- are working on adding support for a larger variety of devices.
-
-
-
${chip} `;
- const chipElementTableData = (value) => `${chipElement(value)} `;
- const chipElementEstimation = (count) =>
- `${count} `;
- items.forEach((item) => {
- chipString += `
-
- ${chipElementTableData(item.value)}
- ${chipElementEstimation(item.count)}
-
- `;
- });
- $(".wl-property-panel__frequent-items").html(chipString);
- if (infType === "non-discrete") {
- $propertyPanelTitle.html("Histogram data:");
- $propertyPanelProfileName.html("Bin values");
- } else if (infType === "discrete") {
- $propertyPanelTitle.html("Frequent items:");
- $propertyPanelProfileName.html("Counts");
- }
+ const chipElement = (chip) => `${chip} `;
+ const chipElementTableData = (value) => `${chipElement(value)} `;
+
+ const chartBoxElement = (chartTitle, chart) => `
+
+
+
${chartTitle}
+
${chart}
+
+
Data Distribution Chart
+
+ ${colorsForDistingushingCharts("#369BAC", "Current")}
+ ${colorsForDistingushingCharts("#2683C9", "Reference")}
+
+ `
+ $(".clickable-test-feature-body").html(`
+ ${chartBoxElement(colorsForDistingushingChartHTMLElement, getDoubleHistogramChart)}
+ `);
+ }
+
+ function categoricalDriftChart(getBarChart, getPositiveNegative) {
+ $(".clickable-test-feature-body").html(`
+ ${chartBoxElement('Bar Chart
', getBarChart)}
+ ${chartBoxElement('Difference Bar Chart
', getPositiveNegative)}
+ `);
+ }
+
+ function sortWithIndeces(toSort) {
+ for (let i = 0; i < toSort.length; i++) {
+ toSort[i] = [toSort[i], i];
+ }
+ toSort.sort((left, right) => {
+ return right[0] < left[0] ? -1 : 1;
+ });
+ toSort.sortIndices = [];
+ for (let j = 0; j < toSort.length; j++) {
+ toSort.sortIndices.push(toSort[j][1]);
+ toSort[j] = toSort[j][0];
+ }
+ return toSort;
+ }
+
+ const getProfileCharts = (key, getDoubleHistogramChart, getBarChart, getPositiveNegative) => {
+ if (jsonData.columns[key].numberSummary.isDiscrete) {
+ $("#page-button").text("Categorical Data")
+ categoricalDriftChart(getBarChart, getPositiveNegative)
} else {
- let chipString = "";
- $propertyPanelTitle.html("");
- const chipElement = (chip) => `${chip} `;
- const chipElementTableData = (value) => `${chipElement(value)} `;
- const chipElementEstimation = (count) =>
- `${count} `;
- const columns = [];
- let itemsLength = items[0].length;
- selectedProfiles.forEach((selected, index) => {
- if (selected !== null && items[parseInt(selected)].length > 0 && items !== "undefined") {
- columns[index] = [];
- items[parseInt(selected)].forEach((item, i) => {
- columns[index][i] = "";
-
- if (index === 0) {
- columns[index][i] += `${chipElementTableData(item.value)}`;
- }
- let value = Number.isInteger(Number(item.count)) ? item.count : roundToThreeDecimals(item.count);
- columns[index][i] += `${chipElementEstimation(value)}`;
- });
+ $("#page-button").text("Numerical Data")
+ numericalDriftChart(getDoubleHistogramChart)
+ }
+ }
+
+ function openReferencePropertyPanel(referenceItems, items, profileItems, key, chart, getDoubleHistogramChart, getBarChart, getPositiveNegative) {
+ const chartInfoItem = (drift, driftName) => `
+
+
${drift}
+
${driftName}
+
+
+
+ items
+
+ ${frequentItemString}
+
+
+
+ reference profile items
+
+ ${referenceFrequentItemString}
+
+
${count} `;
+ items.forEach((item) => {
+ chipString += `
+
+ ${chipElementTableData(item.value)}
+ ${chipElementEstimation(item.count)}
+
+ `;
+ });
+ $(".wl-property-panel__frequent-items").html(chipString);
+ if (infType === "non-discrete") {
+ $propertyPanelTitle.html("Histogram data:");
+ $propertyPanelProfileName.html("Bin values");
+ } else if (infType === "discrete") {
+ $propertyPanelTitle.html("Frequent items:");
+ $propertyPanelProfileName.html("Counts");
+ }
+
+ $(".wl-property-panel").addClass("wl-property-panel--open");
+ $(".wl-table-wrap").addClass("wl-table-wrap--narrow");
+ }
+
+ function openPropertyPanel(referenceItems, items, infType, feature) {
+ let getGraph = null,
+ getPropertyPanelGraph = null,
+ getDoubleHistogramChart,
+ getBarChart,
+ getPositiveNegative,
+ currentWidth = 600,
+ propertyPanelGraph = true;
+
+
+ if (referencePropertyPanelData[feature[0]][0]) {
+ getDoubleHistogramChart = generateDoubleHistogramChart(currentWidth, feature[1].chartData[0], feature[1].chartData[1])
+ getBarChart = generateBarChart(currentWidth, feature[1].chartData[0], feature[1].chartData[1])
+ getPositiveNegative = generatePositiveNegativeChart(currentWidth, feature[1].chartData[0], feature[1].chartData[1])
+ items = referencePropertyPanelData[feature[0]][0]
+ getGraph = getGraphHtml(feature[1].chartData[1], 50, 280, 0, true, propertyPanelGraph)
+ }
+
+ getPropertyPanelGraph = getPropertyPanelGraphHtml(jsonData.columns[feature[0]], feature[0])
+
+ if (jsonData) {
+ if (items.length > 0 && items !== "undefined") {
+ if (referencePropertyPanelData[feature[0]][0]) {
+ openReferencePropertyPanel(
+ referenceItems,
+ items,
+ referenceJsonData.columns[feature[0]],
+ feature[0],
+ getGraph,
+ getDoubleHistogramChart,
+ getBarChart,
+ getPositiveNegative
+ )
} else {
- columns[index] = [];
- for (let i = 0; i < itemsLength; i++) {
- columns[index][i] = "`;
- for (let j = 0; j < selectedProfiles.length; j++) {
- chipString += columns[j][i];
- }
- chipString += ` `;
}
-
- $(".wl-property-panel").addClass("wl-property-panel--open");
- $(".wl-table-wrap").addClass("wl-table-wrap--narrow");
- $(".wl-property-panel__frequent-items").html(chipString);
}
}
@@ -235,81 +701,106 @@
$(".wl-property-panel__frequent-items").html("");
}
- function roundToThreeDecimals(str) {
- let floatNum = parseFloat(str);
- let floatRounded = floatNum.toFixed(3);
+ function getPropertyPanelGraphHtml (column) {
+ let chartString = "";
+ const freqData = [];
+ const histData = [];
+
+ const freqChart = (chart) =>
+ `Frequent Items Data
${chart}
${
chartData.length > 0
- ? getGraphHtml(chartData, profileSelected)
+ ? getGraphHtml(feature[1].chartData[0])
: 'No data to show the chart '
}
`;
+ if (feature[1].chartData[1] && feature[1].chartData[1].length > 0) {
+ referenceTempChartDataString+= `${
+ chartData.length > 0
+ ? getGraphHtml(feature[1].chartData[1], ...[,,], 1, true)
+ : 'No data to show the chart '
+ }
`;
+ }
+ }
+ });
+
+ let diffFromRef = "";
+ feature[1].frequentItemsElemString.forEach((frequentItemElemString, index) => {
+ if (selectedProfiles.includes(String(index))) {
+ diffFromRef += ` ${Math.floor(Math.random() * 10)}
`;
}
});
let freaquentItemsElmString = "";
feature[1].frequentItemsElemString.forEach((frequentItemElemString, index) => {
if (selectedProfiles.includes(String(index))) {
- freaquentItemsElmString += ` ${frequentItemElemString}
`;
+ freaquentItemsElmString += ` ${frequentItemElemString}
`;
}
});
let inferredTypeString = "";
@@ -418,9 +909,15 @@
${feature[0]}
- ` +
- tempChartDataString +
- `
` +
+
+
` +
+ tempChartDataString +
+ referenceTempChartDataString +
+ `
+
` +
+ checkCurrentProfile(``, diffFromRef) +
+ `
` +
freaquentItemsElmString +
`
` +
inferredTypeString +
@@ -450,26 +947,17 @@
quantilesMaxString +
`
`,
);
- if (!checkJSONValidityForMultiProfile(jsonData)) {
- const $tableRowButton = $(`View details `);
- $tableRowButton.on(
- "click",
- openPropertyPanel.bind(this, propertyPanelData[feature[0]][0], feature[1].inferredType[0].toLowerCase()),
- );
- $tableRow.find(".wl-table-cell__title-wrap").append($tableRowButton);
- } else {
+ if (jsonData) {
const $tableRowButton = $(`View details `);
-
- let secondType = feature[1].inferredType[selectedProfiles[1]] || null;
- if (typeof secondType === "string") secondType = secondType.toLowerCase();
$tableRowButton.on(
"click",
openPropertyPanel.bind(
this,
- propertyPanelData[feature[0]],
- feature[1].inferredType[selectedProfiles[0]].toLowerCase(),
- secondType,
+ referencePropertyPanelData[feature[0]][0],
+ propertyPanelData[feature[0]][0],
+ feature[1].inferredType[0].toLowerCase(),
+ feature
),
);
$tableRow.find(".wl-table-cell__title-wrap").append($tableRowButton);
@@ -478,10 +966,22 @@
$tableBody.append($tableRow);
$sidebarFeatureNameList.append(
- `${feature[0]}
+
+
+
+
${feature[0]}
+
+ `,
);
});
- if (!checkJSONValidityForMultiProfile(jsonData)) {
+
+ if (jsonData) {
const countDiscrete = Object.values(numOfProfilesBasedOnType).reduce(
(acc, feature) => (acc += feature.discrete.length),
0,
@@ -498,7 +998,10 @@
$featureCountDiscrete.html(countDiscrete);
$featureCountNonDiscrete.html(countNonDiscrete);
$featureCountUnknown.html(countUnknown);
- $selectedProfile.html(formatLabelDate(+dataForRead.properties.dataTimestamp));
+ $selectedProfile.html(formatLabelDate(+dataForRead.properties[0].dataTimestamp));
+ if (checkCurrentProfile(false, true) && dataForRead.properties[1]) {
+ $selectedReferenceProfile.html(formatLabelDate(+dataForRead.properties[1].dataTimestamp))
+ }
}
}
@@ -511,7 +1014,6 @@
});
for (let i = 0; i < tableBodyChildrens.length; i++) {
- const name = tableBodyChildrens[i].dataset.featureName.toLowerCase();
const type = tableBodyChildrens[i].dataset.inferredType.toLowerCase();
if (isActiveInferredType[type]) {
@@ -520,7 +1022,6 @@
}
}
for (let i = 0; i < featureListChildren.length; i++) {
- const name = featureListChildren[i].dataset.featureName.toLowerCase();
const type = featureListChildren[i].dataset.inferredType.toLowerCase();
if (isActiveInferredType[type]) {
@@ -530,96 +1031,45 @@
$featureCount.html(featureCount);
}
- function renderProfileDropdown(id) {
- $profileDropdown.html("");
-
- $profileDropdown.append(
- `Select your profile `,
- );
- for (let i = 0; i < profiles.length; i++) {
- if (selectedProfiles.includes(String(i))) {
- const option = `${profiles[i].label} `;
- $profileDropdown.append(option);
- } else {
- const option = `${profiles[i].label} `;
- $profileDropdown.append(option);
- }
+ function getDataForRead(dataForRead, jsonData, referneceData) {
+ if (!dataForRead.properties) {
+ dataForRead.properties = [];
}
- }
+ dataForRead.properties.push(jsonData.properties);
- function reRenderProfileDropdown(id) {
- $(`#sidebar-content-multi-profile-dropdown-${id}`).html("");
- let selected = "";
- if (selectedProfiles[id] === null || selectedProfiles.length - 1 < id) {
- selected = "selected";
- }
- $(`#sidebar-content-multi-profile-dropdown-${id}`).append(
- `Select your profile `,
- );
- for (let i = 0; i < profiles.length; i++) {
- if (selectedProfiles.includes(String(i))) {
- let selected = "";
- if (selectedProfiles[id] === String(i)) {
- selected = "selected";
- }
- const option = `${profiles[i].label} `;
- $(`#sidebar-content-multi-profile-dropdown-${id}`).append(option);
- } else {
- const option = `${profiles[i].label} `;
- $(`#sidebar-content-multi-profile-dropdown-${id}`).append(option);
+ Object.entries(jsonData.columns).forEach((feature) => {
+ let tempFeatureName = feature[0];
+ if (!dataForRead.columns) {
+ dataForRead.columns = [];
}
- }
- }
-
- function handleProfileChange(event) {
- handleClosePropertyPanel();
- $tableContent.removeClass("d-none");
- $("#table-content-wrapper").removeClass("d-none");
- const value = event.target.value;
- const id = event.target.dataset;
- selectedProfiles[parseInt(id.id)] = value;
- let addButton = $(`#add-profile-sidebar-button-${id.id}`);
- addButton.attr("disabled", false);
- for (let i = 0; i < selectedProfiles.length; i++) {
- reRenderProfileDropdown(i);
- }
- $selectOptionFirstTime.addClass("d-none");
- $tableBody.html("");
- updateHtmlElementValues();
- renderList();
- }
-
- function mapProfileDataToReadData(jsonData, dataForRead) {
- if (checkJSONValidityForMultiProfile(jsonData)) {
- Object.entries(jsonData).forEach((profile) => {
- if (!dataForRead.properties) {
- dataForRead.properties = [];
- }
- dataForRead.properties.push(profile[1].properties);
-
- Object.entries(profile[1].columns).forEach((feature) => {
- let tempFeatureName = feature[0];
- if (!dataForRead.columns) {
- dataForRead.columns = [];
+ if (!dataForRead.columns[tempFeatureName]) dataForRead.columns[tempFeatureName] = [];
+ dataForRead.columns[tempFeatureName].push(feature[1]);
+ if (
+ referneceData &&
+ referneceData.columns[tempFeatureName].numberSummary
+ ) {
+ dataForRead.properties.push(referenceJsonData.properties)
+ const {
+ numberSummary,
+ frequentItems,
+ ...referenceDataForRead
+ } = dataForRead.columns[tempFeatureName][0]
+ dataForRead.columns[tempFeatureName].push({
+ ...referenceDataForRead,
+ referenceNumberSummary: referneceData.columns[tempFeatureName].numberSummary,
+ referenceFrequentItems: referneceData.columns[tempFeatureName].frequentItems
+ })
}
- if (!dataForRead.columns[tempFeatureName]) dataForRead.columns[tempFeatureName] = [];
- dataForRead.columns[tempFeatureName].push(feature[1]);
- });
- });
- } else {
- if (!dataForRead.properties) {
- dataForRead.properties = [];
- }
- dataForRead.properties.push(jsonData.properties);
+ });
+ return dataForRead
+ }
- Object.entries(jsonData.columns).forEach((feature) => {
- let tempFeatureName = feature[0];
- if (!dataForRead.columns) {
- dataForRead.columns = [];
- }
- if (!dataForRead.columns[tempFeatureName]) dataForRead.columns[tempFeatureName] = [];
- dataForRead.columns[tempFeatureName].push(feature[1]);
- });
+ function mapProfileDataToReadData(jsonData, dataForRead, referneceData) {
+ try {
+ dataForRead = getDataForRead(dataForRead, jsonData, referneceData)
+ } catch (e) {
+ alert("You selected wrong reference profile. Please select again.")
+ removeReferenceProfile()
}
makeFeatureDataForAllProfilesToShowOnTable(
featureDataForTableForAllProfiles,
@@ -628,6 +1078,7 @@
);
}
+
function makeFeatureDataForAllProfilesToShowOnTable(
featureDataForTableForAllProfiles,
dataForRead,
@@ -652,6 +1103,7 @@
mean: [],
stddev: [],
chartData: [],
+ refereneChartData: [],
frequentItemsElemString: [],
};
numOfProfilesBasedOnType[feature[0]] = {
@@ -660,6 +1112,7 @@
unknown: [],
};
propertyPanelData[feature[0]] = [];
+ referencePropertyPanelData[feature[0]] = [];
}
let iteration = 0;
feature[1].forEach((tempFeatureValues) => {
@@ -717,6 +1170,14 @@
axisX: index,
});
});
+ if (tempFeatureValues.referenceFrequentItems) {
+ tempFeatureValues.referenceFrequentItems.items.forEach((item, index) => {
+ featureDataForTableForAllProfiles[feature[0]].chartData[1].push({
+ axisY: item.estimate,
+ axisX: index,
+ });
+ });
+ }
// Frequent item chips / bedge
propertyPanelData[feature[0]][iteration] = tempFeatureValues.frequentItems.items.reduce((acc, item) => {
@@ -737,7 +1198,7 @@
// Chart
if (tempFeatureValues.numberSummary) {
// Histogram chips / bedge
- propertyPanelData[feature[0]][iteration] = tempFeatureValues.numberSummary.histogram.counts.reduce(
+ propertyPanelData[feature[0]][0] = tempFeatureValues.numberSummary.histogram.counts.reduce(
(acc, value, index) => {
acc.push({
value: value,
@@ -749,7 +1210,7 @@
);
tempFeatureValues.numberSummary.histogram.counts.slice(0, 30).forEach((count, index) => {
- featureDataForTableForAllProfiles[feature[0]].chartData[iteration].push({
+ featureDataForTableForAllProfiles[feature[0]].chartData[0].push({
axisY: count,
axisX: index,
});
@@ -763,7 +1224,37 @@
featureDataForTableForAllProfiles[feature[0]].chartData[iteration] = [];
propertyPanelData[feature[0]] = [];
}
+ if (tempFeatureValues.referenceNumberSummary) {
+ referencePropertyPanelData[feature[0]][0] = tempFeatureValues.referenceNumberSummary.histogram.counts.reduce(
+ (acc, value, index) => {
+ acc.push({
+ value: value,
+ count: tempFeatureValues.referenceNumberSummary.histogram.bins[index],
+ });
+ return acc;
+ },
+ [],
+ );
+
+ tempFeatureValues.referenceNumberSummary.histogram.counts.slice(0, 30).forEach((count, index) => {
+ featureDataForTableForAllProfiles[feature[0]].chartData[1].push({
+ axisY: count,
+ axisX: index,
+ });
+ });
+ }
}
+ if(tempFeatureValues.referenceFrequentItems &&
+ featureDataForTableForAllProfiles[feature[0]].inferredType[0] === "Discrete"
+ ){
+ featureDataForTableForAllProfiles[feature[0]].chartData[1] = []
+ tempFeatureValues.referenceFrequentItems.items.forEach((item, index) => {
+ featureDataForTableForAllProfiles[feature[0]].chartData[1].push({
+ axisY: item.estimate,
+ axisX: index,
+ });
+ });
+ }
iteration += 1;
});
});
@@ -774,14 +1265,6 @@
$tableMessage.find("p").html(message);
}
- function collectProfilesFromJSON(data) {
- return Object.keys(data).map((profile, i) => ({
- label: profile,
- value: profile,
- index: i,
- }));
- }
-
function compareArrays(array, target) {
if (array.length !== target.length) return false;
@@ -793,7 +1276,6 @@
return false;
}
}
-
return true;
}
@@ -827,15 +1309,39 @@
$tableMessage.removeClass("d-none");
$sidebarContent.addClass("d-none");
$tableContent.addClass("d-none");
+ $compareProfile.removeClass("d-none");
+ }
+
+ function addFileName() {
+ const fileName = $fileInput.val().split('\\').pop();
+ if (fileName) {
+ $featureFileName.html(fileName.split('.json'))
+ }
+ }
+
+ function removeReferenceProfile() {
+ referenceJsonData = undefined
+ dataForRead = {};
+ featureDataForTableForAllProfiles = {};
+ numOfProfilesBasedOnType = {};
+
+ $(".reference-table-head").addClass("d-none")
+
+ mapProfileDataToReadData(jsonData, dataForRead, referenceJsonData);
+ $tableBody.html("");
+ updateHtmlElementValues();
+ renderList();
+ $(".wl-selected-profile").addClass("d-none")
+ $(".wl-compare-profile").removeClass("d-none")
}
- function loadFile() {
- profiles = [];
+ function loadFile(inputId, jsonForm) {
isActiveInferredType = {};
$featureFilterInput.html("");
$sidebarFeatureNameList.html("");
$tableBody.html("");
$(".form-check-input").prop("checked", true);
+ addFileName()
if (typeof window.FileReader !== "function") {
updateTableMessage(MESSAGES.error.fileAPINotSupported);
return;
@@ -845,7 +1351,7 @@
numOfProfilesBasedOnType = {};
handleClosePropertyPanel();
- const input = document.getElementById("file-input");
+ const input = document.getElementById(inputId);
if (!input) {
updateTableMessage(MESSAGES.error.noInputElementFound);
} else if (!input.files) {
@@ -855,53 +1361,44 @@
} else {
const file = input.files[0];
const fr = new FileReader();
- fr.onload = receivedText;
+ fr.onload = (e) => receivedText(e, jsonForm, inputId);
fr.readAsText(file);
}
}
- function receivedText(e) {
+ function receivedText(e, jsonForm, inputId) {
const lines = e.target.result;
- jsonData = JSON.parse(lines);
- if (Object.keys(dataForRead).length !== 0) {
- dataForRead = {};
- }
- mapProfileDataToReadData(jsonData, dataForRead);
- if (selectedProfiles.length > 0) {
- selectedProfiles = [];
- }
-
- if (checkJSONValidityForMultiProfile(jsonData)) {
- for (let i = 0; i < 2; i++) {
- let addButton = $(`#add-profile-sidebar-button-${i}`);
- addButton.removeClass("d-none");
- addButton.attr("disabled", true);
+ let data = JSON.parse(lines);
+ if (inputId === "file-input") {
+ if (checkJSONValidityForMultiProfile(data)) {
+ jsonData = Object.values(data)[0]
+ } else {
+ jsonData = data;
}
-
- profiles = collectProfilesFromJSON(jsonData);
- renderProfileDropdown();
+ removeReferenceProfile()
+ $(".reference-table-head").addClass("d-none")
+ $(".wl-selected-profile").addClass("d-none")
+ $compareProfile.removeClass("d-none");
+ } else {
+ if (checkJSONValidityForMultiProfile(data)) {
+ referenceJsonData = Object.values(data)[0]
+ } else {
+ referenceJsonData = data;
+ }
+ $(".reference-table-head").removeClass("d-none")
+ $(".wl-selected-profile").removeClass("d-none")
+ $(".wl-compare-profile").addClass("d-none")
+ }
+ mapProfileDataToReadData(jsonData, dataForRead, referenceJsonData);
+ if (data) {
$(".compare-select").addClass("d-none");
- $tableMessage.removeClass("d-none");
- $tableContent.addClass("d-none");
- $("#table-content-wrapper").addClass("d-none");
- updateTableMessage(MESSAGES.error.noProfileSelected);
$multiProfileWrap.removeClass("d-none");
- $singleProfileWrap.addClass("d-none");
- $filterOptions.addClass("d-none");
- $sidebarContent.removeClass("d-none");
- renderList();
- $jsonForm.trigger("reset");
- } else if (checkJSONValidityForSingleProfile(jsonData)) {
- $(".compare-select").addClass("d-none");
- $multiProfileWrap.addClass("d-none");
- $singleProfileWrap.removeClass("d-none");
- $filterOptions.removeClass("d-none");
selectedProfiles[0] = "0";
$tableBody.html("");
updateHtmlElementValues();
renderList();
showDataVisibility();
- $jsonForm.trigger("reset");
+ jsonForm.trigger("reset");
} else {
$tableBody.html("");
updateTableMessage(MESSAGES.error.invalidJSONFile);
@@ -910,60 +1407,87 @@
}
}
- for (let i = 0; i < 2; i++) {
- let addButton = $(`#add-profile-sidebar-button-${i}`);
-
- addButton.on("click", function () {
- reRenderProfileDropdown(i + 1);
- if ($(`#sidebar-content-multi-profile-dropdown-${i} option:selected`).text() !== "Select your profile") {
- addButton.addClass("d-none");
-
- $(`#add-profile-wrap-${i + 1}`).addClass("button-remove-select");
- $(`#add-profile-wrap-${i + 1}`).removeClass("d-none");
- $(`#remove-button-${i + 1}`).removeClass("d-none");
- } else {
+ function checkedBoxes() {
+ const item = Object.values($boxes).find(
+ function(value) {
+ return $('#' + $(value)[0].id).is(":checked")
}
- });
- }
+ );
- $removeButton.on("click", function () {
- handleClosePropertyPanel();
- $(`#add-profile-wrap-1`).addClass("d-none");
- $("#sidebar-content-multi-profile-dropdown-1").val("none");
- $removeButton.addClass("d-none");
- let addButton = $(`#add-profile-sidebar-button-1`);
- addButton.attr("disabled", true);
- $(`#add-profile-wrap-0`).removeClass("button-remove-select");
- $(`#remove-button-0`).addClass("d-none");
- $(`#add-profile-sidebar-button-0`).removeClass("d-none");
- selectedProfiles[1] = null;
- for (let i = 0; i < selectedProfiles.length; i++) {
- reRenderProfileDropdown(i);
+ if (item) {
+ $notifCircleContainer.removeClass("d-none")
}
- $tableBody.html("");
- updateHtmlElementValues();
- renderList();
+ }
+
+ $removeReferenceProfileButton.on("click", removeReferenceProfile)
+
+ $(document).on("click", ".page-button", function(e) {
+ const $pagesButtons = $(".page-button"),
+ $pagesButtonIndex = $pagesButtons.index(e.target),
+ $pagesButton = $(".page-button")[$pagesButtonIndex]
+
+ $pagesButtons.removeClass("activ-pages-button")
+ $($pagesButton).addClass("activ-pages-button")
+ })
+
+ $(window).ready(function() {
+ sidebarContentHeight()
+ })
+
+ $("#property-panel-close-icon").on("click", function (e) {
+ const $clickableTestFeatureWrap = $(".clickable-test-feature-wrap")
+
+ $tableContent.removeClass("d-none")
+ $clickableTestFeatureWrap.addClass("d-none")
});
- $removeButton2.on("click", function () {
- handleClosePropertyPanel();
- $(`#add-profile-wrap-2`).addClass("d-none");
- $("#sidebar-content-multi-profile-dropdown-1").val("none");
- $removeButton2.addClass("d-none");
- $(`#add-profile-wrap-1`).removeClass("button-remove-select");
- $(`#add-profile-sidebar-button-1`).removeClass("d-none");
- selectedProfiles[2] = null;
- for (let i = 0; i < selectedProfiles.length; i++) {
- reRenderProfileDropdown(i);
+ $dropdownArrowIcon.on("click", function () {
+ const filterClass = $filterOptions.attr("class");
+
+ if (filterClass.indexOf("d-none") > 0) {
+ $notifCircleContainer.addClass("d-none")
+ $filterOptions.removeClass("d-none");
+ $("#dropdown-container").css("background-color", '#FFF')
+ $dropdownArrowIcon.css("transform","rotate(180deg)")
+ } else {
+ $filterOptions.addClass("d-none");
+ checkedBoxes()
+ $("#dropdown-container").css("background-color", 'none')
+ $dropdownArrowIcon.css("transform","rotate(0)")
}
- $tableBody.html("");
- updateHtmlElementValues();
- renderList();
});
+
+ $closeIcon.on("click", function () {
+ $signUpText.addClass("d-none");
+ sidebarContentHeight()
+ $(".open-sign-up-text-notif-container").removeClass("d-none")
+ });
+
+ $openSignUpText.on("click", function () {
+ $signUpText.removeClass("d-none");
+ sidebarContentHeight()
+ $(".open-sign-up-text-notif-container").addClass("d-none")
+ });
+
+
+ $(document).on("click", ".js-list-group-item span", function (e) {
+ const listItem = $("li>span"),
+ listItemIndex = listItem.index(e.target),
+ $listItemDot = $(".wl_list-item-dot")[listItemIndex],
+ $arrowIcon = $(".wl_arrow-icon")[listItemIndex]
+
+ listItem.css("padding-left", "15px")
+ $(".wl_list-item-dot").removeClass("d-none")
+ $(".wl_arrow-icon").addClass("d-none")
+ $($arrowIcon).removeClass("d-none")
+ $($listItemDot).addClass("d-none")
+ $(listItem[listItemIndex]).css("padding-left", "8px")
+ });
+
// Bind event listeners
- $fileInput.on("change", loadFile);
+ $fileInput.on("change", () => loadFile("file-input", $jsonForm));
+ $referencefileInput.on("change", () => loadFile("reference-file-input", $referenceJsonForm));
- $profileDropdown.on("change", handleProfileChange);
$(document).on("click", ".js-list-group-item span", scrollToFeatureName);
$featureSearch.on(
"input",
diff --git a/src/whylogs/viewer/templates/index-hbs-cdn-all-in-for-jupyter-notebook.html b/src/whylogs/viewer/templates/index-hbs-cdn-all-in-for-jupyter-notebook.html
new file mode 100644
index 0000000000..d339ac3d6a
--- /dev/null
+++ b/src/whylogs/viewer/templates/index-hbs-cdn-all-in-for-jupyter-notebook.html
@@ -0,0 +1,1352 @@
+
+
+
+ Profile Viewer | whylogs
+
+ Profile Viewer | whylogs
+
+ Profile Viewer | whylogs
+
+ Profile Viewer | whylogs
+
+ Profile Viewer | whylogs
+
+ Profile Viewer | whylogs
+
+ Profile Viewer | whylogs
+
+ Profile Viewer | whylogs
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
+
Clickable test feature
+
+
+
+
+
+
+
+
+ Numerical Data
+
+
+ Frequent items
+
+
+
+
+
+
+
+
+
+
+
+
Compare profile to reference profile, such as training data
+
+
+
+
+
+
+
+
Diff from ref
+
Frequent items
+
Inf. feature type
+
Total count
+
Null fraction
+
Est. unique values
+
Data type
+
Data type count
+
Mean
+
Std. dev
+
Min
+
First quantile
+
Median
+
Third quantile
+
Max
+
+
+
+
+
+
+
Please select and load a JSON profile from your local drive.
+
+
+
+
+
+
+
+
Hold on! :)
+
+ It looks like your current screen size or device is not yet supported by the WhyLabs Sandbox. The Sandbox is
+ best experienced on a desktop computer. Please try maximizing this window or switching to another device. We
+ are working on adding support for a larger variety of devices.
+
+
+
"""
+ return HTML(iframe)
+
+ def set_profiles(self, target_profile: DatasetProfile = None, reference_profile: DatasetProfile = None):
+ self._target_profile = target_profile
+ self._reference_profile = reference_profile
+ if self._target_profile:
+ self._target_profile_json = message_to_json(self._target_profile.to_summary())
+ if self._reference_profile:
+ self._reference_profile_json = message_to_json(self._reference_profile.to_summary())
+
+ def summary_drift_report(self, preferred_cell_height=None):
+ reference_profile = add_drift_val_to_ref_profile_json(self._target_profile, self._reference_profile, json.loads(self._reference_profile_json))
+ template = self.__get_compiled_template(self.SUMMARY_REPORT_TEMPLATE_NAME)
+ profiles_summary = {"profile_from_whylogs": self._target_profile_json}
+ if self._reference_profile:
+ profiles_summary["reference_profile_from_whylogs"] = json.dumps(reference_profile)
+ return self.__display_rendered_template(template(profiles_summary), self.SUMMARY_REPORT_TEMPLATE_NAME, preferred_cell_height)
+
+ def double_histogram(self, feature_names, preferred_cell_height=None):
+ return self.__display_feature_chart(feature_names, self.DOUBLE_HISTOGRAM_TEMPLATE_NAME, preferred_cell_height)
+
+ def distribution_chart(self, feature_names, preferred_cell_height=None):
+ return self.__display_feature_chart(feature_names, self.DISTRIBUTION_CHART_TEMPLATE_NAME, preferred_cell_height)
+
+ def difference_distribution_chart(self, feature_names, preferred_cell_height=None):
+ return self.__display_feature_chart(feature_names, self.DIFFERENCED_CHART_TEMPLATE_NAME, preferred_cell_height)
+
+ def feature_statistics(self, feature_name, profile="reference", preferred_cell_height=None):
+ template = self.__get_compiled_template(self.FEATURE_STATISTICS_TEMPLATE_NAME)
+ if self._reference_profile and profile.lower() == "reference":
+ selected_profile_json = self._reference_profile_json
+ selected_profile = self._reference_profile.columns
+ else:
+ selected_profile_json = self._target_profile_json
+ selected_profile = self._target_profile.columns
+ if selected_profile.get(feature_name).schema_tracker.to_summary().inferred_type.type in numerical_types:
+ rendered_template = template(
+ {
+ "profile_feature_statistics_from_whylogs": json.dumps(
+ add_feature_statistics(selected_profile.get(feature_name), selected_profile_json, feature_name)
+ )
+ }
+ )
+ return self.__display_rendered_template(rendered_template, self.FEATURE_STATISTICS_TEMPLATE_NAME, preferred_cell_height)
+ else:
+ logger.warning("Quantile and descriptive statistics can be calculated for numerical features only!")
+ return None
+
+ def constraints_report(self, constraints, preferred_cell_height=None):
+ template = self.__get_compiled_template(self.CONSTRAINTS_REPORT_TEMPLATE_NAME)
+ rendered_template = template({"constraints_report": json.dumps(constraints.report())})
+ return self.__display_rendered_template(rendered_template, self.CONSTRAINTS_REPORT_TEMPLATE_NAME, preferred_cell_height)
+
+ def download(self, html, preferred_path=None, html_file_name=None):
+ if not html_file_name:
+ if self._reference_profile:
+ html_file_name = self._reference_profile.dataset_timestamp
+ else:
+ html_file_name = self._target_profile.dataset_timestamp
+ if preferred_path:
+ path = os.path.join(os.path.expanduser(preferred_path), str(html_file_name) + ".html")
+ else:
+ path = os.path.join(os.pardir, "html_reports", str(html_file_name) + ".html")
+ full_path = os.path.abspath(path)
+ with open(full_path, "w") as saved_html:
+ saved_html.write(html.data)
+ saved_html.close()
+ return None
diff --git a/src/whylogs/viz/utils/__init__.py b/src/whylogs/viz/utils/__init__.py
new file mode 100644
index 0000000000..e69de29bb2
diff --git a/src/whylogs/viz/utils/profile_viz_calculations.py b/src/whylogs/viz/utils/profile_viz_calculations.py
new file mode 100644
index 0000000000..7d1f1d5a9d
--- /dev/null
+++ b/src/whylogs/viz/utils/profile_viz_calculations.py
@@ -0,0 +1,180 @@
+import json
+
+from whylogs.core.summaryconverters import (
+ compute_chi_squared_test_p_value,
+ ks_test_compute_p_value,
+ single_quantile_from_sketch,
+)
+from whylogs.proto import InferredType, ReferenceDistributionDiscreteMessage
+
+categorical_types = (InferredType.Type.INTEGRAL, InferredType.Type.STRING, InferredType.Type.BOOLEAN)
+
+
+def __calculate_variance(profile_jsons, feature_name):
+ """
+ Calculates variance for single feature
+
+ Parameters
+ ----------
+ profile_jsons: Profile summary serialized json
+ feature_name: Name of feature
+
+ Returns
+ -------
+ variance : Calculated variance for feature
+ """
+ feature = profile_jsons.get("columns").get(feature_name)
+ variance = feature.get("numberSummary").get("stddev") ** 2 if feature.get("numberSummary") is not None else 0
+ return variance
+
+
+def __calculate_coefficient_of_variation(profile_jsons, feature_name):
+ """
+ Calculates coefficient of variation for single feature
+
+ Parameters
+ ----------
+ profile_jsons: Profile summary serialized json
+ feature_name: Name of feature
+
+ Returns
+ -------
+ coefficient_of_variation : Calculated coefficient of variation for feature
+ """
+ feature = profile_jsons.get("columns").get(feature_name)
+ coefficient_of_variation = (
+ feature.get("numberSummary").get("stddev") / feature.get("numberSummary").get("mean") if feature.get("numberSummary") is not None else 0
+ )
+ return coefficient_of_variation
+
+
+def __calculate_sum(profile_jsons, feature_name):
+ """
+ Calculates sum for single feature
+
+ Parameters
+ ----------
+ profile_jsons: Profile summary serialized json
+ feature_name: Name of feature
+
+ Returns
+ -------
+ coefficient_of_variation : Calculated sum for feature
+ """
+ feature = profile_jsons.get("columns").get(feature_name)
+ feature_number_summary = feature.get("numberSummary")
+ if feature_number_summary:
+ sum = feature_number_summary.get("mean") * int(feature.get("counters").get("count"))
+ else:
+ sum = 0
+ return sum
+
+
+def __calculate_quantile_statistics(feature, profile_jsons, feature_name):
+ """
+ Calculates sum for single feature
+
+ Parameters
+ ----------
+ profile_jsons: Profile summary serialized json
+ feature_name: Name of feature
+
+ Returns
+ -------
+ coefficient_of_variation : Calculated sum for feature
+ """
+ quantile_statistics = {}
+ feature_number_summary = profile_jsons.get("columns").get(feature_name).get("numberSummary")
+ if feature.number_tracker and feature.number_tracker.histogram.get_n() > 0:
+ kll_sketch = feature.number_tracker.histogram
+ quantile_statistics["fifth_percentile"] = single_quantile_from_sketch(kll_sketch, quantile=0.05).quantile
+ quantile_statistics["q1"] = single_quantile_from_sketch(kll_sketch, quantile=0.25).quantile
+ quantile_statistics["median"] = single_quantile_from_sketch(kll_sketch, quantile=0.5).quantile
+ quantile_statistics["q3"] = single_quantile_from_sketch(kll_sketch, quantile=0.75).quantile
+ quantile_statistics["ninety_fifth_percentile"] = single_quantile_from_sketch(kll_sketch, quantile=0.95).quantile
+ quantile_statistics["range"] = feature_number_summary.get("max") - feature_number_summary.get("min")
+ quantile_statistics["iqr"] = quantile_statistics["q3"] - quantile_statistics["q1"]
+ return quantile_statistics
+
+
+def add_drift_val_to_ref_profile_json(target_profile, reference_profile, reference_profile_json):
+ """
+ Calculates drift value for reference profile based on profile type and inserts that data into reference profile
+
+ Parameters
+ ----------
+ target_profile: Target profile
+ reference_profile: Reference profile
+ reference_profile_json: Reference profile summary serialized json
+
+ Returns
+ -------
+ reference_profile_json : Reference profile summary serialized json with drift value for every feature
+ """
+ observations = 0
+ missing_cells = 0
+ total_count = 0
+ for target_col_name in target_profile.columns.keys():
+ target_col = target_profile.columns[target_col_name]
+ observations += target_col.counters.to_protobuf().count
+ null_count = target_col.to_summary().counters.null_count.value
+ missing_cells += null_count if null_count else 0
+ total_count += target_col.to_summary().counters.count
+
+ if target_col_name in reference_profile.columns:
+ ref_col = reference_profile.columns[target_col_name]
+ target_type = target_col.schema_tracker.to_summary().inferred_type.type
+ ref_type = ref_col.schema_tracker.to_summary().inferred_type.type
+ if all([type == InferredType.FRACTIONAL for type in (ref_type, target_type)]):
+ target_kll_sketch = target_col.number_tracker.histogram
+ reference_kll_sketch = ref_col.number_tracker.histogram
+ ks_p_value = ks_test_compute_p_value(target_kll_sketch, reference_kll_sketch)
+ reference_profile_json["columns"][target_col_name]["drift_from_ref"] = ks_p_value.ks_test
+ elif all([type in categorical_types for type in (ref_type, target_type)]) and ref_type == target_type:
+ target_frequent_items_sketch = target_col.frequent_items
+ reference_frequent_items_sketch = ref_col.frequent_items
+ if any([msg.to_summary() is None for msg in (target_frequent_items_sketch, reference_frequent_items_sketch)]):
+ continue
+ target_total_count = target_col.counters.count
+ target_message = ReferenceDistributionDiscreteMessage(frequent_items=target_frequent_items_sketch.to_summary(), total_count=target_total_count)
+ ref_total_count = ref_col.counters.count
+
+ reference_message = ReferenceDistributionDiscreteMessage(
+ frequent_items=reference_frequent_items_sketch.to_summary(), total_count=ref_total_count
+ )
+ chi_squared_p_value = compute_chi_squared_test_p_value(target_message, reference_message)
+ if chi_squared_p_value.chi_squared_test:
+ reference_profile_json["columns"][target_col_name]["drift_from_ref"] = chi_squared_p_value.chi_squared_test
+ else:
+ reference_profile_json["columns"][target_col_name]["drift_from_ref"] = None
+ reference_profile_json["properties"]["observations"] = observations
+ reference_profile_json["properties"]["missing_cells"] = missing_cells
+ reference_profile_json["properties"]["total_count"] = total_count
+ reference_profile_json["properties"]["missing_percentage"] = (missing_cells / total_count) * 100 if total_count else 0
+
+ return reference_profile_json
+
+
+def add_feature_statistics(feature, profile_json, feature_name):
+ """
+ Calculates different values for feature statistics
+
+ Parameters
+ ----------
+ feature:
+ profile_json: Profile summary serialized json
+ feature_name: Name of feature
+
+ Returns
+ -------
+ feature: Feature data with appended values for statistics report
+ """
+ profile_features = json.loads(profile_json)
+ feature_with_statistics = {}
+ feature_with_statistics["properties"] = profile_features.get("properties")
+ feature_with_statistics[feature_name] = profile_features.get("columns").get(feature_name)
+ feature_with_statistics[feature_name]["sum"] = __calculate_sum(profile_features, feature_name)
+ feature_with_statistics[feature_name]["variance"] = __calculate_variance(profile_features, feature_name)
+ feature_with_statistics[feature_name]["coefficient_of_variation"] = __calculate_coefficient_of_variation(profile_features, feature_name)
+ feature_with_statistics[feature_name]["quantile_statistics"] = __calculate_quantile_statistics(feature, profile_features, feature_name)
+ return feature_with_statistics
diff --git a/tests/unit/viz/test_jupyter_notebook_viz.py b/tests/unit/viz/test_jupyter_notebook_viz.py
new file mode 100644
index 0000000000..9a4b06b959
--- /dev/null
+++ b/tests/unit/viz/test_jupyter_notebook_viz.py
@@ -0,0 +1,219 @@
+import datetime
+import os
+
+import numpy as np
+import pandas as pd
+
+from whylogs import get_or_create_session
+from whylogs.core.statistics.constraints import (
+ DatasetConstraints,
+ Op,
+ SummaryConstraint,
+ ValueConstraint,
+ columnPairValuesInSetConstraint,
+ columnsMatchSetConstraint,
+ columnValuesInSetConstraint,
+ columnValuesUniqueWithinRow,
+ sumOfRowValuesOfMultipleColumnsEqualsConstraint,
+)
+from whylogs.viz import NotebookProfileViewer
+
+
+def __generate_target_profile():
+
+ session = get_or_create_session()
+
+ with session.logger("mytestytest", dataset_timestamp=datetime.datetime(2021, 6, 2)) as logger:
+ for _ in range(5):
+ logger.log({"uniform_integers": np.random.randint(0, 50)})
+ logger.log({"nulls": None})
+
+ return logger.profile
+
+
+def __generate_reference_profile():
+
+ session = get_or_create_session()
+
+ with session.logger("mytestytest", dataset_timestamp=datetime.datetime(2021, 6, 2)) as logger:
+ for _ in range(5):
+ logger.log({"uniform_integers": np.random.randint(0, 50)})
+ logger.log({"nulls": None})
+
+ return logger.profile
+
+
+def __generate_categorical_target_profile():
+ session = get_or_create_session()
+ credit_cards = pd.DataFrame(
+ [
+ {"credit_card": "3714-496353-98431"},
+ {"credit_card": "3787 344936 71000"},
+ {"credit_card": "3056 930902 5904"},
+ {"credit_card": "3065 133242 2899"},
+ ]
+ )
+ return session.log_dataframe(credit_cards, "test.data")
+
+
+def __generate_categorical_reference_profile():
+ session = get_or_create_session()
+ credit_cards = pd.DataFrame(
+ [
+ {"credit_card": "6011 1111 1111 1117"},
+ {"credit_card": "6011-0009-9013-9424"},
+ {"credit_card": "3530 1113 3330 0000"},
+ {"credit_card": "3566-0020-2036-0505"},
+ ]
+ )
+ return session.log_dataframe(credit_cards, "test.data")
+
+
+def _get_sample_dataset_constraints():
+ cvisc = columnValuesInSetConstraint(value_set={2, 5, 8})
+ ltc = ValueConstraint(Op.LT, 1)
+
+ min_gt_constraint = SummaryConstraint("min", Op.GT, value=100)
+ max_le_constraint = SummaryConstraint("max", Op.LE, value=5)
+
+ set1 = set(["col1", "col2"])
+ columns_match_constraint = columnsMatchSetConstraint(set1)
+
+ val_set = {(1, 2), (3, 5)}
+ col_set = ["A", "B"]
+ mcv_constraints = [
+ columnValuesUniqueWithinRow(column_A="A", verbose=True),
+ columnPairValuesInSetConstraint(column_A="A", column_B="B", value_set=val_set),
+ sumOfRowValuesOfMultipleColumnsEqualsConstraint(columns=col_set, value=100),
+ ]
+
+ return DatasetConstraints(
+ None,
+ value_constraints={"annual_inc": [cvisc, ltc]},
+ summary_constraints={"annual_inc": [max_le_constraint, min_gt_constraint]},
+ table_shape_constraints=[columns_match_constraint],
+ multi_column_value_constraints=mcv_constraints,
+ )
+
+
+def test_notebook_profile_viewer_set_profiles():
+ target_profile = __generate_target_profile()
+ reference_profile = __generate_reference_profile()
+ viz = NotebookProfileViewer()
+ viz.set_profiles(target_profile=target_profile, reference_profile=reference_profile)
+
+
+def test_summary_drift_report_without_preferred_height():
+ target_profile = __generate_target_profile()
+ reference_profile = __generate_reference_profile()
+ viz = NotebookProfileViewer()
+ viz.set_profiles(target_profile=target_profile, reference_profile=reference_profile)
+ viz.summary_drift_report()
+
+
+def test_summary_drift_report_with_preferred_height():
+ target_profile = __generate_target_profile()
+ reference_profile = __generate_reference_profile()
+ viz = NotebookProfileViewer()
+ viz.set_profiles(target_profile=target_profile, reference_profile=reference_profile)
+ viz.summary_drift_report()
+
+
+def test_feature_statistics_not_passing_profile_type():
+ target_profile = __generate_target_profile()
+ reference_profile = __generate_reference_profile()
+ viz = NotebookProfileViewer()
+ viz.set_profiles(target_profile=target_profile, reference_profile=reference_profile)
+ viz.feature_statistics("uniform_integers")
+
+
+def test_feature_statistics_passing_profile_type():
+ target_profile = __generate_target_profile()
+ reference_profile = __generate_reference_profile()
+ viz = NotebookProfileViewer()
+ viz.set_profiles(target_profile=target_profile, reference_profile=reference_profile)
+ viz.feature_statistics("uniform_integers", "target")
+
+
+def test_feature_statistics_passing_profile_type_and_prefered_height():
+ target_profile = __generate_target_profile()
+ reference_profile = __generate_reference_profile()
+ viz = NotebookProfileViewer()
+ viz.set_profiles(target_profile=target_profile, reference_profile=reference_profile)
+ viz.feature_statistics("uniform_integers", "target", "1000px")
+
+
+def test_download_passing_all_arguments(tmpdir):
+ target_profile = __generate_target_profile()
+ reference_profile = __generate_reference_profile()
+ viz = NotebookProfileViewer()
+ viz.set_profiles(target_profile=target_profile, reference_profile=reference_profile)
+
+ download = viz.download(viz.summary_drift_report(), tmpdir, html_file_name="foo")
+ assert os.path.exists(tmpdir + "/foo.html")
+
+
+def test_constraints_report_without_preferred_height():
+ target_profile = __generate_target_profile()
+ reference_profile = __generate_reference_profile()
+ viz = NotebookProfileViewer()
+ viz.set_profiles(target_profile=target_profile, reference_profile=reference_profile)
+ dc = _get_sample_dataset_constraints()
+ viz.constraints_report(dc)
+
+
+def test_constraints_report_with_preferred_height():
+ target_profile = __generate_target_profile()
+ reference_profile = __generate_reference_profile()
+ viz = NotebookProfileViewer()
+ viz.set_profiles(target_profile=target_profile, reference_profile=reference_profile)
+ dc = _get_sample_dataset_constraints()
+ viz.constraints_report(dc, preferred_cell_height="1000px")
+
+
+def test_double_histogram_without_height():
+ target_profile = __generate_target_profile()
+ reference_profile = __generate_reference_profile()
+ viz = NotebookProfileViewer()
+ viz.set_profiles(target_profile=target_profile, reference_profile=reference_profile)
+ viz.double_histogram("uniform_integers")
+
+
+def test_double_histogram_with_height():
+ target_profile = __generate_target_profile()
+ reference_profile = __generate_reference_profile()
+ viz = NotebookProfileViewer()
+ viz.set_profiles(target_profile=target_profile, reference_profile=reference_profile)
+ viz.double_histogram("uniform_integers", "1000px")
+
+
+def test_distribution_chart_without_height():
+ target_profile = __generate_categorical_target_profile()
+ reference_profile = __generate_categorical_reference_profile()
+ viz = NotebookProfileViewer()
+ viz.set_profiles(target_profile=target_profile, reference_profile=reference_profile)
+ viz.distribution_chart("credit_card")
+
+
+def test_distribution_chart_with_height():
+ target_profile = __generate_categorical_target_profile()
+ reference_profile = __generate_categorical_reference_profile()
+ viz = NotebookProfileViewer()
+ viz.set_profiles(target_profile=target_profile, reference_profile=reference_profile)
+ viz.distribution_chart("credit_card", "1000px")
+
+
+def test_difference_distribution_chart_without_height():
+ target_profile = __generate_categorical_target_profile()
+ reference_profile = __generate_categorical_reference_profile()
+ viz = NotebookProfileViewer()
+ viz.set_profiles(target_profile=target_profile, reference_profile=reference_profile)
+ viz.difference_distribution_chart("credit_card")
+
+
+def test_difference_distribution_chart_with_height():
+ target_profile = __generate_categorical_target_profile()
+ reference_profile = __generate_categorical_reference_profile()
+ viz = NotebookProfileViewer()
+ viz.set_profiles(target_profile=target_profile, reference_profile=reference_profile)
+ viz.difference_distribution_chart("credit_card", "1000px")
+ 
+ 
+ 
+ 
+