add div and glot500 to datasets comparing with.

app.py

@@ -1,6 +1,6 @@
 import pandas as pd
 import streamlit as st
-from utils import df_to_html, render_svg, combine_json_files, render_metadata
+from utils import df_to_html, render_svg, combine_json_files, render_metadata, color_mapping
 
 data = combine_json_files('./languages')
 
@@ -24,10 +24,10 @@ def render_home_table():
     df_data['ISO Code'] = df_data['ISO Code'].astype(str)  # Convert to string
     df_data['Number of Sites'] = df_data.apply(lambda row: '<a href="/?isocode={}&site=True" target="_self">{}</a>'.format(row['ISO Code'], row['Number of Sites']), axis=1)
     df_data['Number of Links'] = df_data.apply(lambda row: '<a href="/?isocode={}&links=True" target="_self">{}</a>'.format(row['ISO Code'], row['Number of Links']), axis=1)
-
+    df_data["Supported by MADLAD-400, flores, and Glot500"] = df_data.apply(lambda row: color_mapping([row["Supported by allenai/MADLAD-400"] + row["Supported by facebook/flores"] +  row["Supported by cis-lmu/Glot500"]]), axis =1)
 
     # Display the table
-    df_data = df_data[['ISO Code', 'Language Name', 'Family', 'Subgrouping', 'Number of Sites', 'Number of Links', 'Number of Speakers', 'Supported by allenai/MADLAD-400 or facebook/flores']]
+    df_data = df_data[['ISO Code', 'Language Name', 'Family', 'Subgrouping', 'Number of Sites', 'Number of Links', 'Number of Speakers', 'Supported by MADLAD-400, flores, and Glot500']]
     st.write(df_to_html(df_data), unsafe_allow_html=True)
 
 

languages/abk_Cyrl.json

@@ -3,7 +3,9 @@
     "Family": "Abkhaz-Adyge",
     "Subgrouping": "Abkhaz-Abaza",
     "Number of Speakers": "190_000",
-    "Supported by allenai/MADLAD-400 or facebook/flores": "🟥",
+    "Supported by allenai/MADLAD-400": 0,
+    "Supported by facebook/flores": 0,
+    "Supported by cis-lmu/Glot500": 1,
     "Sites": [
       {
         "Site Name": "jinepsgazetesi.com",

languages/ady_Cyrl.json

@@ -3,7 +3,9 @@
     "Family": "Abkhaz-Adyge",
     "Subgrouping": "Circassian",
     "Number of Speakers": "610_000",
-    "Supported by allenai/MADLAD-400 or facebook/flores": "🟨",
+    "Supported by allenai/MADLAD-400": 1,
+    "Supported by facebook/flores": 0,
+    "Supported by cis-lmu/Glot500": 0,
     "Sites": [
       {
         "Site Name": "jinepsgazetesi.com",

languages/ahk_Latn.json

@@ -3,7 +3,9 @@
     "Family": "Sino-Tibetan",
     "Subgrouping": "Lolo-Burmese",
     "Number of Speakers": "620_000",
-    "Supported by allenai/MADLAD-400 or facebook/flores": "🟥",
+    "Supported by allenai/MADLAD-400": 1,
+    "Supported by facebook/flores": 0,
+    "Supported by cis-lmu/Glot500": 1,
     "Sites": [
         {
             "Site Name": "akhaliterature.com",

languages/anm_Latn.json

@@ -3,7 +3,9 @@
     "Family": "Sino-Tibetan",
     "Subgrouping": "Kuki-Chin",
     "Number of Speakers": "14_000",
-    "Supported by allenai/MADLAD-400 or facebook/flores": "🟥",
+    "Supported by allenai/MADLAD-400": 0,
+    "Supported by facebook/flores": 0,
+    "Supported by cis-lmu/Glot500": 0,
     "Sites": [
       {
         "Site Name": "apakan.org",

languages/ava_Cyrl.json

@@ -3,7 +3,9 @@
     "Family": "Nakh-Daghestanian",
     "Subgrouping": "Avar–Andic",
     "Number of Speakers": "800_000",
-    "Supported by allenai/MADLAD-400 or facebook/flores": "🟨",
+    "Supported by allenai/MADLAD-400": 1,
+    "Supported by facebook/flores": 0,
+    "Supported by cis-lmu/Glot500": 0,
     "Sites": [
         {
             "Site Name": "radioerkenli.com",

languages/bal_Arab.json

@@ -3,7 +3,9 @@
     "Family": "Indo-European",
     "Subgrouping": "Iranian",
     "Number of Speakers": "8_000_000",
-    "Supported by allenai/MADLAD-400 or facebook/flores": "🟥",
+    "Supported by allenai/MADLAD-400": 0,
+    "Supported by facebook/flores": 0,
+    "Supported by cis-lmu/Glot500": 0,
     "Sites": [
         {
             "Site Name": "sunnionline.us",

languages/bar_Latn.json

@@ -3,7 +3,9 @@
     "Family": "Indo-European",
     "Subgrouping": "Germanic",
     "Number of Speakers": "14_000_000",
-    "Supported by allenai/MADLAD-400 or facebook/flores": "🟥",
+    "Supported by allenai/MADLAD-400": 0,
+    "Supported by facebook/flores": 0,
+    "Supported by cis-lmu/Glot500": 1,
     "Sites": [
         {
             "Site Name": "bar.wikipedia.org",

languages/bqi_Arab.json

@@ -3,7 +3,9 @@
     "Family": "Indo-European",
     "Subgrouping": "Iranian",
     "Number of Speakers": "1_200_000",
-    "Supported by allenai/MADLAD-400 or facebook/flores": "🟥",
+    "Supported by allenai/MADLAD-400": 0,
+    "Supported by facebook/flores": 0,
+    "Supported by cis-lmu/Glot500": 0,
     "Sites": [
         {
             "Site Name": "lurishacademy.org",

languages/brh_Arab.json

@@ -3,7 +3,9 @@
     "Family": "Dravidian",
     "Subgrouping": "",
     "Number of Speakers": "2_800_000",
-    "Supported by allenai/MADLAD-400 or facebook/flores": "🟥",
+    "Supported by allenai/MADLAD-400": 0,
+    "Supported by facebook/flores": 0,
+    "Supported by cis-lmu/Glot500": 0,
     "Sites": [
       {
         "Site Name": "talarbrahui",

languages/che_Cyrl.json

@@ -3,7 +3,9 @@
     "Family": "Nakh-Daghestanian",
     "Subgrouping": "Nakh",
     "Number of Speakers": "1_800_000",
-    "Supported by allenai/MADLAD-400 or facebook/flores": "🟨",
+    "Supported by allenai/MADLAD-400": 1,
+    "Supported by facebook/flores": 0,
+    "Supported by cis-lmu/Glot500": 1,
     "Sites": [
       {
         "Site Name": "jinepsgazetesi.com",

languages/crl_Cans.json

@@ -3,7 +3,9 @@
     "Family": "Algic",
     "Subgrouping": "Cree-Montagnais-Naskapi",
     "Number of Speakers": "400",
-    "Supported by allenai/MADLAD-400 or facebook/flores": "🟥",
+    "Supported by allenai/MADLAD-400": 0,
+    "Supported by facebook/flores": 0,
+    "Supported by cis-lmu/Glot500": 0,    
     "Sites": [
       {
         "Site Name": "nunavikprotectedareas.com",

languages/div_Thaa.json

@@ -0,0 +1,57 @@
+{
+    "Language Name": "Dhivehi",
+    "Family": "Indo-European",
+    "Subgrouping": "Insular Indo-Aryan",
+    "Number of Speakers": "500_000",
+    "Supported by allenai/MADLAD-400": 1,
+    "Supported by facebook/flores": 0,
+    "Supported by cis-lmu/Glot500": 1,
+    "Sites": [
+      {
+        "Site Name": "hassanhameed.com",
+        "Site URL": "https://www.hassanhameed.com/dhivehi-language",
+        "Category": "linguistics",
+        "Confidence": "🟩",
+        "Info": "confirmed by glotlid",
+        "Possible Parallel Languages": "",
+        "Links": []
+      },
+      {
+        "Site Name": "asim-ahmedasim.blogspot.com",
+        "Site URL": "https://asim-ahmedasim.blogspot.com/",
+        "Category": "blog",
+        "Confidence": "🟩",
+        "Info": "confirmed by glotlid",
+        "Possible Parallel Languages": "",
+        "Links": []
+      },
+      {
+        "Site Name": "thaareekh.blogspot.com",
+        "Site URL": "https://thaareekh.blogspot.com/",
+        "Category": "",
+        "Confidence": "🟩",
+        "Info": "confirmed by glotlid",
+        "Possible Parallel Languages": "",
+        "Links": []
+      },
+      {
+        "Site Name": "ilyashussainibrahim.com",
+        "Site URL": "https://www.ilyashussainibrahim.com/",
+        "Category": "",
+        "Confidence": "🟩",
+        "Info": "confirmed by glotlid",
+        "Possible Parallel Languages": "",
+        "Links": []
+      },
+      {
+        "Site Name": "halinews.com",
+        "Site URL": "https://halinews.com/",
+        "Category": "news",
+        "Confidence": "🟩",
+        "Info": "confirmed by glotlid",
+        "Possible Parallel Languages": "",
+        "Links": []
+      }
+    ]
+  }
+  

languages/enm_Latn.json

@@ -3,7 +3,9 @@
     "Family": "Indo-European",
     "Subgrouping": "Germanic",
     "Number of Speakers": "0",
-    "Supported by allenai/MADLAD-400 or facebook/flores": "🟥",
+    "Supported by allenai/MADLAD-400": 0,
+    "Supported by facebook/flores": 0,
+    "Supported by cis-lmu/Glot500": 1,
     "Sites": [
       {
         "Site Name": "quod.lib.umich.edu/c/cme/AHA2727/",

languages/glk_Arab.json

@@ -3,7 +3,9 @@
     "Family": "Indo-European",
     "Subgrouping": "Iranian",
     "Number of Speakers": "3_000_000",
-    "Supported by allenai/MADLAD-400 or facebook/flores": "🟥",
+    "Supported by allenai/MADLAD-400": 0,
+    "Supported by facebook/flores": 0,
+    "Supported by cis-lmu/Glot500": 1, 
     "Sites": [
       {
         "Site Name": "t.me/gilaki_twitter",

languages/hac_Arab.json

@@ -3,7 +3,9 @@
     "Family": "Indo-European",
     "Subgrouping": "Iranian",
     "Number of Speakers": "300_000",
-    "Supported by allenai/MADLAD-400 or facebook/flores": "🟥",
+    "Supported by allenai/MADLAD-400": 0,
+    "Supported by facebook/flores": 0,
+    "Supported by cis-lmu/Glot500": 0,
     "Sites": [
       {
         "Site Name": "anfsorani.com/هۆرامی",

languages/ike_Cans.json

@@ -3,7 +3,9 @@
     "Family": "Eskimo-Aleut",
     "Subgrouping": "Inuit",
     "Number of Speakers": "42_000",
-    "Supported by allenai/MADLAD-400 or facebook/flores": "🟨",
+    "Supported by allenai/MADLAD-400": 1,
+    "Supported by facebook/flores": 0,
+    "Supported by cis-lmu/Glot500": 1,
     "Sites": [
       {
         "Site Name": "nunavikprotectedareas.com",

languages/khb_Talu.json

@@ -3,7 +3,9 @@
     "Family": "Tai-Kadai",
     "Subgrouping": "Daic",
     "Number of Speakers": "550_000",
-    "Supported by allenai/MADLAD-400 or facebook/flores": "🟥",
+    "Supported by allenai/MADLAD-400": 0,
+    "Supported by facebook/flores": 0,
+    "Supported by cis-lmu/Glot500": 0,
     "Sites": [
       {
         "Site Name": "dw12.com",

languages/lki_Arab.json

@@ -3,7 +3,9 @@
     "Family": "Indo-European",
     "Subgrouping": "Iranian",
     "Number of Speakers": "600_000",
-    "Supported by allenai/MADLAD-400 or facebook/flores": "🟥",
+    "Supported by allenai/MADLAD-400": 0,
+    "Supported by facebook/flores": 0,
+    "Supported by cis-lmu/Glot500": 0,
     "Sites": [
         {
             "Site Name": "lurishacademy.org",

languages/lrc_Arab.json

@@ -3,7 +3,9 @@
     "Family": "Indo-European",
     "Subgrouping": "Iranian",
     "Number of Speakers": "4_000_000",
-    "Supported by allenai/MADLAD-400 or facebook/flores": "🟨",
+    "Supported by allenai/MADLAD-400": 1,
+    "Supported by facebook/flores": 0,
+    "Supported by cis-lmu/Glot500": 0,
     "Sites": [
         {
             "Site Name": "lurishacademy.org",

languages/mns_Cyrl.json

@@ -3,7 +3,9 @@
   "Family": "Uralic",
   "Subgrouping": "Mansic",
   "Number of Speakers": "2_200",
-  "Supported by allenai/MADLAD-400 or facebook/flores": "🟥",
+  "Supported by allenai/MADLAD-400": 0,
+  "Supported by facebook/flores": 0,
+  "Supported by cis-lmu/Glot500": 0,
   "Sites": [
     {
       "Site Name": "khanty-yasang.ru/luima-seripos",

languages/mon_Mong.json

@@ -3,7 +3,9 @@
     "Family": "Mongolic",
     "Subgrouping": "",
     "Number of Speakers": "6_000_000",
-    "Supported by allenai/MADLAD-400 or facebook/flores": "🟥",
+    "Supported by allenai/MADLAD-400": 0,
+    "Supported by facebook/flores": 0,
+    "Supported by cis-lmu/Glot500": 0,
     "Sites": [
         {
             "Site Name": "wikibilig.mn",

languages/nsk_Cans.json

@@ -3,7 +3,9 @@
     "Family": "Algic",
     "Subgrouping": "Cree-Montagnais-Naskapi",
     "Number of Speakers": "1_200",
-    "Supported by allenai/MADLAD-400 or facebook/flores": "🟥",
+    "Supported by allenai/MADLAD-400": 0,
+    "Supported by facebook/flores": 0,
+    "Supported by cis-lmu/Glot500": 0,
     "Sites": [
       {
         "Site Name": "clscnaskapi.org",

languages/oss_Cyrl.json

@@ -3,7 +3,9 @@
     "Family": "Indo-European",
     "Subgrouping": "Iranian",
     "Number of Speakers": "490_000",
-    "Supported by allenai/MADLAD-400 or facebook/flores": "🟨",
+    "Supported by allenai/MADLAD-400": 1,
+    "Supported by facebook/flores": 0,
+    "Supported by cis-lmu/Glot500": 1,
     "Sites": [
       {
         "Site Name": "jinepsgazetesi.com",

languages/sdh_Arab.json

@@ -3,7 +3,9 @@
     "Family": "Indo-European",
     "Subgrouping": "Iranian",
     "Number of Speakers": "1_500_000",
-    "Supported by allenai/MADLAD-400 or facebook/flores": "🟥",
+    "Supported by allenai/MADLAD-400": 0,
+    "Supported by facebook/flores": 0,
+    "Supported by cis-lmu/Glot500": 0,
     "Sites": [
       {
         "Site Name": "shafaq.com/ku",

languages/snk_Latn.json

@@ -3,7 +3,9 @@
     "Family": "Mande",
     "Subgrouping": "Soninke-Bozo",
     "Number of Speakers": "2_300_000",
-    "Supported by allenai/MADLAD-400 or facebook/flores": "🟥",
+    "Supported by allenai/MADLAD-400": 0,
+    "Supported by facebook/flores": 0,
+    "Supported by cis-lmu/Glot500": 0,
     "Sites": [
       {
         "Site Name": "soninkara.com/snk/",

languages/srp_Latn.json

@@ -3,7 +3,9 @@
     "Family": "Indo-European",
     "Subgrouping": "Slavic",
     "Number of Speakers": "12_000_000",
-    "Supported by allenai/MADLAD-400 or facebook/flores": "🟥",
+    "Supported by allenai/MADLAD-400": 0,
+    "Supported by facebook/flores": 0,
+    "Supported by cis-lmu/Glot500": 1,
     "Sites": [
       {
         "Site Name": "turizamiputovanja.com",

languages/syl_Sylo.json

@@ -3,7 +3,9 @@
     "Family": "Indo-European",
     "Subgrouping": "Eastern Indo-Aryan",
     "Number of Speakers": "10_000_000",
-    "Supported by allenai/MADLAD-400 or facebook/flores": "🟥",
+    "Supported by allenai/MADLAD-400": 0,
+    "Supported by facebook/flores": 0,
+    "Supported by cis-lmu/Glot500": 0,
     "Sites": [
         {
             "Site Name": "surmafarorkhobor.com",

languages/uzs_Arab.json

@@ -3,7 +3,9 @@
     "Family": "Turkic",
     "Subgrouping": "Turkestan",
     "Number of Speakers": "4_600_000",
-    "Supported by allenai/MADLAD-400 or facebook/flores": "🟥",
+    "Supported by allenai/MADLAD-400": 0,
+    "Supported by facebook/flores": 0,
+    "Supported by cis-lmu/Glot500": 0,
     "Sites": [
       {
         "Site Name": "trt.net.tr/afghaniuzbek",

languages/xmf_Geor.json

@@ -3,7 +3,9 @@
     "Family": "Kartvelian",
     "Subgrouping": "Zan",
     "Number of Speakers": "345_000",
-    "Supported by allenai/MADLAD-400 or facebook/flores": "🟥",
+    "Supported by allenai/MADLAD-400": 0,
+    "Supported by facebook/flores": 0,
+    "Supported by cis-lmu/Glot500": 1,
     "Sites": [
       {
         "Site Name": "titus.uni-frankfurt.de",

utils.py

@@ -81,4 +81,23 @@ def render_metadata():
         <a href="https://arxiv.org/abs/xxxx.xxxxx"><img alt="arXiv" src="https://img.shields.io/badge/arXiv-xxxx.xxxxx-b31b1b.svg"></a>
         </p>"""
     c = st.container()
-    c.write(html, unsafe_allow_html=True)
+    c.write(html, unsafe_allow_html=True)
+
+
+
+
+# Define the function for conditional coloring
+def color_mapping(listx):
+
+    bins = len(listx)
+    mean = sum(listx) / (1.0 * bins)
+
+    if mean == 0:
+        return "🟥"
+    elif mean < 1.0/bins + 0.01:
+        return "🟧"
+    elif mean < 2.0/bins + 0.01:
+        return "🟨"
+    else:
+        return "🟩"
+