Merge branch 'graph_improvment' into mkdocs-work

setup.py

@@ -11,7 +11,7 @@
 setup(
     name="SynRD",
     version="0.1",
-    python_requires='== 3.7.*', # Unfortunately, need specific python version for rpy2
+    python_requires='== 3.9.*', # Unfortunately, need specific python version for rpy2
     description="Benchmark for differentially private synthetic data.",
     long_description=long_description,
     author="Lucas Rosenblatt",

web/formatted_data.json

@@ -1 +1 @@
-{"data": [[0.0, 1.0, 1.0, 1.0, 1.0, 0.0, 0.6799999999999998, 1.0, 1.0, 1.0, 1.0, 1.0], [1.0, 1.0, 1.0, 0.6399999999999999, 0.5999999999999999, 0.5999999999999999, 0.6399999999999999, 1.0, 1.0, 1.0, 1.0, 1.0]], "columns": ["iverson22football-0", "iverson22football-1", "iverson22football-2", "iverson22football-3", "iverson22football-4", "iverson22football-5", "iverson22football-6", "lee2021ability-0", "lee2021ability-1", "lee2021ability-2", "lee2021ability-3", "lee2021ability-4"], "index": ["privbayes - 0.37", "privbayes - 1.0"]}
+{"data": [[0.0, 1.0, 1.0, 1.0, 1.0, 0.0, 0.6799999999999998, 1.0, 1.0, 1.0, 1.0, 1.0], [1.0, 1.0, 1.0, 0.6399999999999999, 0.5999999999999999, 0.5999999999999999, 0.6399999999999999, 1.0, 1.0, 1.0, 1.0, 1.0]], "columns": ["iverson22football-0", "iverson22football-1", "iverson22football-2", "iverson22football-3", "iverson22football-4", "iverson22football-5", "iverson22football-6", "lee2021ability-0", "lee2021ability-1", "lee2021ability-2", "lee2021ability-3", "lee2021ability-4"], "index": ["privbayes - 0.37", "privbayes - 1.0"], "findings": [["The sample was, on average, 38 years old<br>at the Wave V assessment.", "During the Wave V assessment, 307<br>(17.4%) men reported being diagnosed<br>with depression at some point in their<br>life, 275 (15.6%) being diagnosed with<br>an anxiety disorder or panic disorder at<br>some point in their life, 211 (12.0%)<br>having received psychological or<br>emotional counseling in the past 12<br>months, 125 (7.1%) reported seriously<br>thinking about suicide in the past year,<br>and 101 (5.8%) reported feeling<br>depressed in the previous week (i.e., \u201ca<br>lot of the time\u201d or \u201cmost of the time or<br>all of the time\u201d over the past 7 days).", "Examining responses the participants<br>gave during the Wave I assessment when<br>they were adolescents, 369 (20.9%)<br>reported playing (or intending to play)<br>football in high school and 952 (54.0%)<br>reported not intending to play football<br>in high school. Of note, 441<br>participants (25% of the sample) did not<br>answer this question and were excluded<br>from analyses pertaining to football<br>participation.", "During Wave I, there were 174 boys<br>(9.9%) who reported undergoing<br>psychological counseling in the past<br>year while in high school. During the<br>Wave V interview, ~24 years later, those<br>individuals who underwent psychological<br>counseling during adolescence were much<br>more likely to report (i) a lifetime<br>history of depression [37.4 vs. 15.3%,<br>\u03c72 ((11)) = 53.17, p &lt; 0.001, OR=<br>3.31, 95% CI = 2.37-4.64],", "(ii) a lifetime history of an anxiety<br>disorder or panic disorder [31.0 vs.<br>14.9%, \u03c72 ((11)) = 41.12, p &lt; 0.001,<br>OR= 2.61, 95% CI = 1.94-3.50],", "(iii) having received psychological<br>counseling in the past 12 months [31.0<br>vs. 12.0%, \u03c72 ((11)) = 36.00, p &lt;<br>0.001, OR= 2.70, 95% CI = 1.98-3.67],", "(iv) suicidal ideation in the past year<br>[12.6 vs. 6.4%, \u03c72(1) = 9.24 , 95%CI =<br>1.29\u20133.44], , p = 0.002, OR = 2.11", "As predicted, a negative correlation was<br>found between perceived low math teacher<br>support and 11thgrade math achievement<br>(r = -0.11).", "Ability self-concepts and parental<br>supportin 9th grade were positively<br>correlated with students'11th grade<br>achievement (r = 0.30 and r = 0.12<br>respectively).", "A strong positive correlation was found<br>between9th and 11th grade math<br>achievement (r = 0.74).", "perceived low math teacher support in<br>11th gradenegatively predicted students'<br>11th grade math achievement(B = -1.51, p<br>&lt; .001) while controlling for<br>students'demographics, 9th grade math<br>achievement score, and mathcourse.", "Second, as shown under Model 2, math<br>ability self-conceptspositively and<br>directly predicted 11th grade math<br>achievement(B = 2.51, p &lt; .001)."], ["The sample was, on average, 38 years old<br>at the Wave V assessment.", "During the Wave V assessment, 307<br>(17.4%) men reported being diagnosed<br>with depression at some point in their<br>life, 275 (15.6%) being diagnosed with<br>an anxiety disorder or panic disorder at<br>some point in their life, 211 (12.0%)<br>having received psychological or<br>emotional counseling in the past 12<br>months, 125 (7.1%) reported seriously<br>thinking about suicide in the past year,<br>and 101 (5.8%) reported feeling<br>depressed in the previous week (i.e., \u201ca<br>lot of the time\u201d or \u201cmost of the time or<br>all of the time\u201d over the past 7 days).", "Examining responses the participants<br>gave during the Wave I assessment when<br>they were adolescents, 369 (20.9%)<br>reported playing (or intending to play)<br>football in high school and 952 (54.0%)<br>reported not intending to play football<br>in high school. Of note, 441<br>participants (25% of the sample) did not<br>answer this question and were excluded<br>from analyses pertaining to football<br>participation.", "During Wave I, there were 174 boys<br>(9.9%) who reported undergoing<br>psychological counseling in the past<br>year while in high school. During the<br>Wave V interview, ~24 years later, those<br>individuals who underwent psychological<br>counseling during adolescence were much<br>more likely to report (i) a lifetime<br>history of depression [37.4 vs. 15.3%,<br>\u03c72 ((11)) = 53.17, p &lt; 0.001, OR=<br>3.31, 95% CI = 2.37-4.64],", "(ii) a lifetime history of an anxiety<br>disorder or panic disorder [31.0 vs.<br>14.9%, \u03c72 ((11)) = 41.12, p &lt; 0.001,<br>OR= 2.61, 95% CI = 1.94-3.50],", "(iii) having received psychological<br>counseling in the past 12 months [31.0<br>vs. 12.0%, \u03c72 ((11)) = 36.00, p &lt;<br>0.001, OR= 2.70, 95% CI = 1.98-3.67],", "(iv) suicidal ideation in the past year<br>[12.6 vs. 6.4%, \u03c72(1) = 9.24 , 95%CI =<br>1.29\u20133.44], , p = 0.002, OR = 2.11", "As predicted, a negative correlation was<br>found between perceived low math teacher<br>support and 11thgrade math achievement<br>(r = -0.11).", "Ability self-concepts and parental<br>supportin 9th grade were positively<br>correlated with students'11th grade<br>achievement (r = 0.30 and r = 0.12<br>respectively).", "A strong positive correlation was found<br>between9th and 11th grade math<br>achievement (r = 0.74).", "perceived low math teacher support in<br>11th gradenegatively predicted students'<br>11th grade math achievement(B = -1.51, p<br>&lt; .001) while controlling for<br>students'demographics, 9th grade math<br>achievement score, and mathcourse.", "Second, as shown under Model 2, math<br>ability self-conceptspositively and<br>directly predicted 11th grade math<br>achievement(B = 2.51, p &lt; .001)."]]}

web/index.html

@@ -12,14 +12,17 @@
         fetch('formatted_data.json')
   .then(response => response.json())
   .then(data => {
+    console.log(data.columns.length)
+    console.log( Array.from({ length: data.columns.length }, (_, i) => i + 1))
     // Prepare data for Plotly
     const trace = {
       z: data.data,
-      x: Array.from({ length: data.columns.length }, (_, i) => i + 1),
+      x: [data.columns.map(col => col.split('-')[0]), data.columns.map(col => parseInt(col.split('-')[1]) + 1)],
       y: data.index,
       type: 'heatmap',
-      hoverinfo: 'z+x+y',
-      colorscale: [[0, 'black'], [1, 'white']],
+      hovertext: data.findings,
+      hoverinfo: 'text',
+      colorscale: [[0, 'FD7900'], [1, '80FB35']],
       showscale: true,
       xgap: 1,
       ygap: 1,
@@ -58,11 +61,14 @@
     // Layout configuration
     const layout = {
       title: 'Epistemic Parity',
-      xaxis: {
-        tickvals: Array.from({ length: data.columns.length }, (_, i) => i),
-        ticktext: Array.from({ length: data.columns.length }, (_, i) => i + 1),
-      },
-      annotations: annotations,
+      yaxis:{
+        tickangle:270
+      }
+      // xaxis: {
+      //   tickvals: Array.from({ length: data.columns.length }, (_, i) => i),
+      //   ticktext: Array.from({ length: data.columns.length }, (_, i) => i + 1),
+      // },
+      // annotations: annotations,
     };
 
     const config = {

web/preprocess.ipynb

@@ -2,12 +2,32 @@
  "cells": [
   {
    "cell_type": "code",
-   "execution_count": 1,
+   "execution_count": 27,
    "metadata": {},
-   "outputs": [],
+   "outputs": [
+    {
+     "name": "stdout",
+     "output_type": "stream",
+     "text": [
+      "{'iverson22football': <class 'SynRD.papers.iverson22football.Iverson22Football'>, 'lee2021ability': <class 'SynRD.papers.lee2021ability.Lee2021Ability'>}\n",
+      "Index([0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11], dtype='int64')\n",
+      "Index([12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23], dtype='int64')\n",
+      "{'data': [[0.0, 1.0, 1.0, 1.0, 1.0, 0.0, 0.6799999999999998, 1.0, 1.0, 1.0, 1.0, 1.0], [1.0, 1.0, 1.0, 0.6399999999999999, 0.5999999999999999, 0.5999999999999999, 0.6399999999999999, 1.0, 1.0, 1.0, 1.0, 1.0]], 'columns': ['iverson22football-0', 'iverson22football-1', 'iverson22football-2', 'iverson22football-3', 'iverson22football-4', 'iverson22football-5', 'iverson22football-6', 'lee2021ability-0', 'lee2021ability-1', 'lee2021ability-2', 'lee2021ability-3', 'lee2021ability-4'], 'index': ['privbayes - 0.37', 'privbayes - 1.0'], 'findings': [['The sample was, on average, 38 years old<br>at the Wave V assessment.', 'During the Wave V assessment, 307<br>(17.4%) men reported being diagnosed<br>with depression at some point in their<br>life, 275 (15.6%) being diagnosed with<br>an anxiety disorder or panic disorder at<br>some point in their life, 211 (12.0%)<br>having received psychological or<br>emotional counseling in the past 12<br>months, 125 (7.1%) reported seriously<br>thinking about suicide in the past year,<br>and 101 (5.8%) reported feeling<br>depressed in the previous week (i.e., “a<br>lot of the time” or “most of the time or<br>all of the time” over the past 7 days).', 'Examining responses the participants<br>gave during the Wave I assessment when<br>they were adolescents, 369 (20.9%)<br>reported playing (or intending to play)<br>football in high school and 952 (54.0%)<br>reported not intending to play football<br>in high school. Of note, 441<br>participants (25% of the sample) did not<br>answer this question and were excluded<br>from analyses pertaining to football<br>participation.', 'During Wave I, there were 174 boys<br>(9.9%) who reported undergoing<br>psychological counseling in the past<br>year while in high school. During the<br>Wave V interview, ~24 years later, those<br>individuals who underwent psychological<br>counseling during adolescence were much<br>more likely to report (i) a lifetime<br>history of depression [37.4 vs. 15.3%,<br>χ2 ((11)) = 53.17, p &lt; 0.001, OR=<br>3.31, 95% CI = 2.37-4.64],', '(ii) a lifetime history of an anxiety<br>disorder or panic disorder [31.0 vs.<br>14.9%, χ2 ((11)) = 41.12, p &lt; 0.001,<br>OR= 2.61, 95% CI = 1.94-3.50],', '(iii) having received psychological<br>counseling in the past 12 months [31.0<br>vs. 12.0%, χ2 ((11)) = 36.00, p &lt;<br>0.001, OR= 2.70, 95% CI = 1.98-3.67],', '(iv) suicidal ideation in the past year<br>[12.6 vs. 6.4%, χ2(1) = 9.24 , 95%CI =<br>1.29–3.44], , p = 0.002, OR = 2.11', 'As predicted, a negative correlation was<br>found between perceived low math teacher<br>support and 11thgrade math achievement<br>(r = -0.11).', \"Ability self-concepts and parental<br>supportin 9th grade were positively<br>correlated with students'11th grade<br>achievement (r = 0.30 and r = 0.12<br>respectively).\", 'A strong positive correlation was found<br>between9th and 11th grade math<br>achievement (r = 0.74).', \"perceived low math teacher support in<br>11th gradenegatively predicted students'<br>11th grade math achievement(B = -1.51, p<br>&lt; .001) while controlling for<br>students'demographics, 9th grade math<br>achievement score, and mathcourse.\", 'Second, as shown under Model 2, math<br>ability self-conceptspositively and<br>directly predicted 11th grade math<br>achievement(B = 2.51, p &lt; .001).'], ['The sample was, on average, 38 years old<br>at the Wave V assessment.', 'During the Wave V assessment, 307<br>(17.4%) men reported being diagnosed<br>with depression at some point in their<br>life, 275 (15.6%) being diagnosed with<br>an anxiety disorder or panic disorder at<br>some point in their life, 211 (12.0%)<br>having received psychological or<br>emotional counseling in the past 12<br>months, 125 (7.1%) reported seriously<br>thinking about suicide in the past year,<br>and 101 (5.8%) reported feeling<br>depressed in the previous week (i.e., “a<br>lot of the time” or “most of the time or<br>all of the time” over the past 7 days).', 'Examining responses the participants<br>gave during the Wave I assessment when<br>they were adolescents, 369 (20.9%)<br>reported playing (or intending to play)<br>football in high school and 952 (54.0%)<br>reported not intending to play football<br>in high school. Of note, 441<br>participants (25% of the sample) did not<br>answer this question and were excluded<br>from analyses pertaining to football<br>participation.', 'During Wave I, there were 174 boys<br>(9.9%) who reported undergoing<br>psychological counseling in the past<br>year while in high school. During the<br>Wave V interview, ~24 years later, those<br>individuals who underwent psychological<br>counseling during adolescence were much<br>more likely to report (i) a lifetime<br>history of depression [37.4 vs. 15.3%,<br>χ2 ((11)) = 53.17, p &lt; 0.001, OR=<br>3.31, 95% CI = 2.37-4.64],', '(ii) a lifetime history of an anxiety<br>disorder or panic disorder [31.0 vs.<br>14.9%, χ2 ((11)) = 41.12, p &lt; 0.001,<br>OR= 2.61, 95% CI = 1.94-3.50],', '(iii) having received psychological<br>counseling in the past 12 months [31.0<br>vs. 12.0%, χ2 ((11)) = 36.00, p &lt;<br>0.001, OR= 2.70, 95% CI = 1.98-3.67],', '(iv) suicidal ideation in the past year<br>[12.6 vs. 6.4%, χ2(1) = 9.24 , 95%CI =<br>1.29–3.44], , p = 0.002, OR = 2.11', 'As predicted, a negative correlation was<br>found between perceived low math teacher<br>support and 11thgrade math achievement<br>(r = -0.11).', \"Ability self-concepts and parental<br>supportin 9th grade were positively<br>correlated with students'11th grade<br>achievement (r = 0.30 and r = 0.12<br>respectively).\", 'A strong positive correlation was found<br>between9th and 11th grade math<br>achievement (r = 0.74).', \"perceived low math teacher support in<br>11th gradenegatively predicted students'<br>11th grade math achievement(B = -1.51, p<br>&lt; .001) while controlling for<br>students'demographics, 9th grade math<br>achievement score, and mathcourse.\", 'Second, as shown under Model 2, math<br>ability self-conceptspositively and<br>directly predicted 11th grade math<br>achievement(B = 2.51, p &lt; .001).']]}\n"
+     ]
+    }
+   ],
    "source": [
     "import pandas as pd\n",
     "import json\n",
+    "from SynRD.papers import Iverson22Football, Lee2021Ability\n",
+    "import pandas as pd\n",
+    "import textwrap\n",
+    "\n",
+    "# needed to create the instance of publication class\n",
+    "silly_df = pd.DataFrame.from_dict({\"silly\":[1]})\n",
+    "\n",
+    "names_to_papers = {Iverson22Football.DEFAULT_PAPER_ATTRIBUTES['id']:Iverson22Football,\n",
+    "                    Lee2021Ability.DEFAULT_PAPER_ATTRIBUTES['id']: Lee2021Ability}\n",
     "\n",
     "df = pd.read_csv('data.csv')\n",
     "\n",
@@ -17,23 +37,36 @@
     "data = []\n",
     "columns = []\n",
     "index = []\n",
+    "findings = []\n",
+    "\n",
+    "# wrap hovertext function, a line has at most width characters long.\n",
+    "def wrap_hovertext(hovertext, width=40):\n",
+    "    wrapped_hovertext = [\"<br>\".join(textwrap.wrap(text,width=width)) for text in hovertext]\n",
+    "    return wrapped_hovertext\n",
     "\n",
     "for (synthesizer, epsilon), group in grouped:\n",
     "    # sort based on 'paper' and 'finding_number'\n",
     "    group = group.sort_values(['paper', 'finding_number'])\n",
     "    \n",
     "    if not columns:\n",
     "        columns = list(group['paper'] + '-' + group['finding_number'].astype(str))\n",
-    "    \n",
+    "    findings_in_group = []\n",
+    "    for i in group.index:\n",
+    "        paper_name = group['paper'][i]\n",
+    "        paper_class=names_to_papers[paper_name]\n",
+    "        finding = paper_class(silly_df).FINDINGS[group['finding_number'][i]].text.replace('\\n', ' ').replace(2*' ',\"\")\n",
+    "        findings_in_group.append(finding)\n",
+    "\n",
+    "    findings.append(wrap_hovertext(findings_in_group))\n",
     "    index.append(f\"{synthesizer} - {epsilon}\")\n",
     "    data.append(group['parity'].tolist())\n",
     "\n",
     "result = {\n",
     "    'data': data,\n",
     "    'columns': columns,\n",
-    "    'index': index\n",
+    "    'index': index,\n",
+    "    'findings': findings\n",
     "}\n",
-    "\n",
     "with open('formatted_data.json', 'w') as f:\n",
     "    json.dump(result, f)"
    ]
@@ -294,7 +327,7 @@
    "name": "python",
    "nbconvert_exporter": "python",
    "pygments_lexer": "ipython3",
-   "version": "3.7.16"
+   "version": "3.9.18"
   },
   "orig_nbformat": 4,
   "vscode": {