update

Pipfile

@@ -0,0 +1,11 @@
+[[source]]
+url = "https://pypi.org/simple"
+verify_ssl = true
+name = "pypi"
+
+[packages]
+
+[dev-packages]
+
+[requires]
+python_version = "3.9"

Pipfile.lock

@@ -0,0 +1,20 @@
+{
+    "_meta": {
+        "hash": {
+            "sha256": "a36a5392bb1e8bbc06bfaa0761e52593cf2d83b486696bf54667ba8da616c839"
+        },
+        "pipfile-spec": 6,
+        "requires": {
+            "python_version": "3.9"
+        },
+        "sources": [
+            {
+                "name": "pypi",
+                "url": "https://pypi.org/simple",
+                "verify_ssl": true
+            }
+        ]
+    },
+    "default": {},
+    "develop": {}
+}

app.py

@@ -1,40 +1,26 @@
 ## LIBRARIES ###
 ## Data
-import numpy as np
 import pandas as pd
-import torch
-import json
-from tqdm import tqdm
-from math import floor
-from datasets import load_dataset
-from collections import defaultdict
-from transformers import AutoTokenizer
 pd.options.display.float_format = '${:,.2f}'.format
 
 # Analysis
 
 # App & Visualization
 import streamlit as st
-from bokeh.models import CustomJS, ColumnDataSource, HoverTool, BoxSelectTool, Callback, Select, TextInput, DataTable, TableColumn
-from bokeh.events import  SelectionGeometry
-from bokeh.plotting import figure, output_file, show
+from bokeh.models import CustomJS, ColumnDataSource, TextInput, DataTable, TableColumn
+from bokeh.plotting import figure
 from bokeh.transform import factor_cmap
 from bokeh.palettes import Category20c_20
 from bokeh.layouts import column, row
 
 # utils
-from random import sample
 
 def datasets_explorer_viz(df):
     s = ColumnDataSource(df)
-    text_input = TextInput(value="", title="Search")
-    text_input.js_on_change("value", CustomJS(code="""
-        console.log('text_input: value=' + this.value, this.toString())
-    """))
     TOOLTIPS= [("dataset_id", "@dataset_id"), ("task", "@task")]
     color = factor_cmap('task', palette=Category20c_20, factors=df['task'].unique()) 
     p = figure(plot_width=1000, plot_height=1000, tools="hover,wheel_zoom,pan,box_select", title="Dataset explorer", tooltips=TOOLTIPS, toolbar_location="above")
-    p.scatter('x', 'y', size=3, source=s, alpha=0.8,marker='circle',fill_color = color, line_color=color, legend_field = 'task')
+    p.scatter('x', 'y', size=500, source=s, alpha=0.8,marker='circle',fill_color = color, line_color=color, legend_field = 'task')
     p.legend.location = "bottom_right"
     #p.legend.click_policy="mute"
     p.legend.label_text_font_size="8pt"
@@ -51,15 +37,11 @@ def datasets_explorer_viz(df):
             const inds = cb_obj.indices;
             const tableData = table_source.data;
             const umapData = umap_source.data;
-            
-            //tableData['x'] = []
-            //tableData['y'] = []
+
             tableData['task'] = []
             tableData['dataset_id'] = []
 
             for (let i = 0; i < inds.length; i++) {
-                // tableData['x'].push(umapData['x'][inds[i]])
-                // tableData['y'].push(umapData['y'][inds[i]])
                 tableData['task'].push(umapData['task'][inds[i]])
                 tableData['dataset_id'].push(umapData['dataset_id'][inds[i]])
             }
@@ -67,7 +49,21 @@ def datasets_explorer_viz(df):
             table_source.change.emit();
     """
     ))
-    show(row(column(text_input,p), data_table))
+    p.selection = ColumnDataSource(data=dict())
+    #text_input.on_change("value_input", 
+    text_input.value.js_on_change('value', CustomJS(args=dict(plot_source=s, text_input=text_input), code="""
+        const indices = [];
+        const plot_data = plot_source.data;
+        for (var i = 0; i < plot_data['dataset_id'].length(); i++) {
+            console.log(plot_data['dataset_id'][i]);
+            if (plot_data['dataset_id'][i] == text_input.value || plot_data['task'][i] == text_input.value) {
+                indices.push(i)
+            }
+        }
+        plot_source.change.emit()
+    """))
+    text_input = TextInput(value="", title="Search")
+    st.bokeh_chart(row(column(text_input,p), data_table))
 
 
 if __name__ == "__main__":

requirements.txt

@@ -1,285 +1,12 @@
-# This file may be used to create an environment using:
-# $ conda create --name <env> --file <this file>
-# platform: osx-arm64
-bokeh=2.4.2=py39hca03da5_0
-boto3=1.21.32=pyhd3eb1b0_0
-botocore=1.24.32=pyhd3eb1b0_0
-bottleneck=1.3.4=py39heec5a64_0
-brotli=1.0.7=hc377ac9_0
-brotlipy=0.7.0=py39h1a28f6b_1002
-brunsli=0.1=hc377ac9_1
-bzip2=1.0.8=h620ffc9_4
-c-ares=1.18.1=h1a28f6b_0
-c-blosc2=2.0.4=h0095615_1
-ca-certificates=2021.10.8=h4653dfc_0
-cachetools=4.2.2=pyhd3eb1b0_0
-certifi=2021.10.8=py39h2804cbe_2
-cffi=1.15.0=py39h22df2f2_1
-cfitsio=4.0.0=h99351b2_0
-charls=2.2.0=hc377ac9_0
-charset-normalizer=2.0.4=pyhd3eb1b0_0
-click=8.0.4=py39hca03da5_0
-cloudpickle=2.0.0=pyhd3eb1b0_0
-colorcet=3.0.0=py39hca03da5_0
-cryptography=37.0.1=py39h834c97f_0
-cycler=0.11.0=pyhd3eb1b0_0
-cytoolz=0.11.0=py39h1a28f6b_0
-dask=2022.2.1=pyhd3eb1b0_0
-dask-core=2022.2.1=pyhd3eb1b0_0
-dataclasses=0.8=pyh6d0b6a4_7
-datasets=2.0.0=py_0
-datasets-sql=0.1.1=pypi_0
-datashader=0.13.0=pyhd3eb1b0_1
-datashape=0.5.4=py39hca03da5_1
-debugpy=1.5.1=py39hc377ac9_0
-decorator=5.1.1=pyhd3eb1b0_0
-defusedxml=0.7.1=pyhd3eb1b0_0
-dill=0.3.4=pyhd3eb1b0_0
-distributed=2022.2.1=pyhd3eb1b0_0
-docopt=0.6.2=pypi_0
-duckdb=0.3.4=pypi_0
-entrypoints=0.4=py39hca03da5_0
-executing=0.8.3=pyhd3eb1b0_0
-filelock=3.6.0=pyhd3eb1b0_0
-fonttools=4.31.2=pypi_0
-freetype=2.11.0=h1192e45_0
-frozenlist=1.2.0=py39h1a28f6b_0
-fsspec=2022.2.0=pyhd3eb1b0_0
-future=0.18.2=py39hca03da5_1
-fuzzywuzzy=0.18.0=pypi_0
-gflags=2.2.2=hc377ac9_0
-gh=2.10.1=h75b854d_0
-giflib=5.2.1=h1a28f6b_0
-gitdb=4.0.7=pyhd3eb1b0_0
-gitpython=3.1.18=pyhd3eb1b0_1
-glog=0.5.0=hc377ac9_0
-grpc-cpp=1.42.0=hedfbb7c_1
-heapdict=1.0.1=pyhd3eb1b0_0
-holoviews=1.14.8=pyhd3eb1b0_0
-htmlmin=0.1.12=pypi_0
-huggingface-hub=0.6.0=pypi_0
-idna=3.3=pyhd3eb1b0_0
-imagecodecs=2021.11.20=py39hcb02aed_1
-imagehash=4.2.1=pypi_0
-imageio=2.9.0=pyhd3eb1b0_0
-importlib-metadata=4.11.3=py39hca03da5_0
-importlib_metadata=4.11.3=hd3eb1b0_0
-ipykernel=6.9.1=py39hca03da5_0
-ipython=8.3.0=py39hca03da5_0
-ipython_genutils=0.2.0=pyhd3eb1b0_1
-ipywidgets=7.6.5=pyhd3eb1b0_1
-jbig=2.1=h1a28f6b_0
-jedi=0.18.1=py39hca03da5_1
-jellyfish=0.9.0=pypi_0
-jinja2=3.0.3=pyhd3eb1b0_0
-jmespath=0.10.0=pyhd3eb1b0_0
-joblib=1.0.1=pypi_0
-jpeg=9e=h1a28f6b_0
-jsonlines=3.0.0=pypi_0
-jsonschema=4.4.0=py39hca03da5_0
-jupyter=1.0.0=pypi_0
-jupyter-console=6.4.3=pypi_0
-jupyter_client=7.2.2=py39hca03da5_0
-jupyter_core=4.10.0=py39hca03da5_0
-jupyterlab_pygments=0.1.2=py_0
-jupyterlab_widgets=1.0.0=pyhd3eb1b0_1
-jxrlib=1.1=h1a28f6b_2
-kaleido=0.2.1=pypi_0
-kiwisolver=1.4.2=pypi_0
-krb5=1.19.2=h3b8d789_0
-lcms2=2.12=hba8e193_0
-lerc=3.0=hc377ac9_0
-libaec=1.0.6=hbdafb3b_0
-libblas=3.9.0=14_osxarm64_openblas
-libbrotlicommon=1.0.9=h1c322ee_7
-libbrotlidec=1.0.9=h1c322ee_7
-libbrotlienc=1.0.9=h1c322ee_7
-libcblas=3.9.0=14_osxarm64_openblas
-libcurl=7.82.0=hc6d1d07_0
-libcxx=12.0.0=hf6beb65_1
-libdeflate=1.8=h1a28f6b_5
-libedit=3.1.20210910=h1a28f6b_0
-libev=4.33=h1a28f6b_1
-libevent=2.1.10=hbae9a57_4
-libffi=3.4.2=hc377ac9_2
-libgfortran=5.0.0=11_1_0_h6a59814_26
-libgfortran5=11.1.0=h6a59814_26
-liblapack=3.9.0=14_osxarm64_openblas
-liblapacke=3.9.0=14_osxarm64_openblas
-libllvm11=11.1.0=h12f7ac0_4
-libnghttp2=1.46.0=h95c9599_0
-libopenblas=0.3.20=openmp_h2209c59_0
-libpng=1.6.37=hb8d0fd4_0
-libprotobuf=3.19.1=h98b2900_0
-libsodium=1.0.18=h1a28f6b_0
-libssh2=1.10.0=hf27765b_0
-libthrift=0.15.0=h28a9c34_1
-libtiff=4.3.0=h74060c4_2
-libutf8proc=2.6.1=h1a28f6b_0
-libwebp=1.2.2=h68602c7_0
-libwebp-base=1.2.2=h1a28f6b_0
-libzlib=1.2.11=h90dfc92_1014
-libzopfli=1.0.3=hc377ac9_0
-llvm-openmp=14.0.3=hd125106_0
-llvmlite=0.38.0=py39h98b2900_0
-locket=0.2.1=py39hca03da5_2
-lz4-c=1.9.3=hc377ac9_0
-markdown=3.3.4=py39hca03da5_0
-markupsafe=2.0.1=py39h1a28f6b_0
-matplotlib=3.5.1=py39hca03da5_1
-matplotlib-base=3.5.1=py39hc377ac9_1
-matplotlib-inline=0.1.2=pyhd3eb1b0_2
-missingno=0.5.1=pypi_0
-mistune=0.8.4=py39h1a28f6b_1000
-msgpack-python=1.0.3=py39h525c30c_0
-multidict=5.2.0=py39h1a28f6b_2
-multimethod=1.7=pypi_0
-multipledispatch=0.6.0=py39hca03da5_0
-multiprocess=0.70.12.2=py39hb18efdd_2
-munkres=1.1.4=py_0
-nbclient=0.5.13=py39hca03da5_0
-nbconvert=6.4.4=py39hca03da5_0
-nbformat=5.3.0=py39hca03da5_0
-ncurses=6.3=h1a28f6b_2
-nest-asyncio=1.5.5=py39hca03da5_0
-networkx=2.7.1=pyhd3eb1b0_0
-ninja=1.10.2=hca03da5_5
-ninja-base=1.10.2=h525c30c_5
-nltk=3.7=pyhd3eb1b0_0
-notebook=6.4.11=py39hca03da5_0
-numba=0.55.1=py39h9197a36_0
-numexpr=2.8.1=py39h144ceef_0
-numpy=1.21.5=py39h25ab29e_2
-numpy-base=1.21.5=py39h974a1f5_2
-openblas=0.3.20=openmp_h745f6c2_0
-openjpeg=2.4.0=h062765e_1
-openssl=1.1.1o=ha287fd2_0
-orc=1.7.1=hcb6706d_1
-packaging=21.3=pyhd3eb1b0_0
-pandas=1.4.2=py39hc377ac9_0
-pandas-profiling=3.1.0=pypi_0
-pandocfilters=1.5.0=pyhd3eb1b0_0
-panel=0.13.0=py39hca03da5_0
-param=1.12.0=pyhd3eb1b0_0
-parquet-cpp=1.5.1=h34088ae_4
-parso=0.8.3=pyhd3eb1b0_0
-partd=1.2.0=pyhd3eb1b0_1
-pexpect=4.8.0=pyhd3eb1b0_3
-phik=0.12.2=pypi_0
-pickleshare=0.7.5=pyhd3eb1b0_1003
-pillow=9.1.0=pypi_0
-pip=21.2.4=py39hca03da5_0
-pipreqs=0.4.11=pypi_0
-plotly=5.6.0=pyhd3eb1b0_0
-progressbar=2.5=pypi_0
-prometheus_client=0.13.1=pyhd3eb1b0_0
-prompt-toolkit=3.0.20=pyhd3eb1b0_0
-protobuf=3.20.0=pypi_0
-psutil=5.8.0=py39h1a28f6b_1
-ptyprocess=0.7.0=pyhd3eb1b0_2
-pure_eval=0.2.2=pyhd3eb1b0_0
-pyahocorasick=1.4.4=pypi_0
-pyarrow=6.0.1=py39hd3b58d7_5_cpu
-pyasn1=0.4.8=pypi_0
-pycodestyle=2.8.0=pyhd3eb1b0_0
-pycparser=2.21=pyhd3eb1b0_0
-pyct=0.4.6=py39hca03da5_0
-pydantic=1.9.0=pypi_0
-pydeck=0.7.1=pyh6c4a22f_0
-pygments=2.11.2=pyhd3eb1b0_0
-pympler=1.0.1=pypi_0
-pynndescent=0.5.4=pyhd3eb1b0_0
-pyopenssl=22.0.0=pyhd3eb1b0_0
-pyparsing=3.0.4=pyhd3eb1b0_0
-pyrsistent=0.18.0=py39h1a28f6b_0
-pysocks=1.7.1=py39hca03da5_0
-python=3.9.12=hfc7342c_1_cpython
-python-dateutil=2.8.2=pyhd3eb1b0_0
-python-dotenv=0.19.2=pypi_0
-python-fastjsonschema=2.15.3=pyhd8ed1ab_0
-python-xxhash=3.0.0=py39hb18efdd_1
-python_abi=3.9=1_cp39
-pytorch=1.10.2=cpu_py39h23cb94c_0
-pytz=2021.3=pyhd3eb1b0_0
-pyviz_comms=2.0.2=pyhd3eb1b0_0
-pywavelets=1.3.0=py39h1a28f6b_0
-pyyaml=6.0=py39h1a28f6b_0
-pyzmq=22.3.0=py39hc377ac9_2
-qtconsole=5.3.0=pypi_0
-qtpy=2.0.1=pypi_0
-re2=2021.11.01=hbdafb3b_0
-readline=8.1.2=h1a28f6b_1
-regex=2022.3.15=py39h1a28f6b_0
-requests=2.27.1=pyhd3eb1b0_0
-responses=0.18.0=pypi_0
-s3transfer=0.5.0=pyhd3eb1b0_0
-sacremoses=0.0.43=pyhd3eb1b0_0
-scikit-image=0.19.2=py39h9197a36_0
-scikit-learn=1.0.2=py39h9197a36_1
-scipy=1.7.3=py39h2f0f56f_0
-seaborn=0.11.2=pyhd3eb1b0_0
-segtok=1.5.11=pypi_0
-semver=2.13.0=pyhd3eb1b0_0
-send2trash=1.8.0=pyhd3eb1b0_1
-sentence-transformers=2.2.0=pyhd8ed1ab_0
-sentencepiece=0.1.95=py39h525c30c_0
-setuptools=61.2.0=py39hca03da5_0
-simplejson=3.17.6=pypi_0
-six=1.16.0=pyhd3eb1b0_1
-smmap=5.0.0=pypi_0
-snappy=1.1.9=hc377ac9_0
-sortedcontainers=2.4.0=pyhd3eb1b0_0
-soupsieve=2.3.2=pypi_0
-sql-metadata=2.5.0=pypi_0
-sqlite=3.38.2=h1058600_0
-sqlparse=0.4.2=pypi_0
-stack_data=0.2.0=pyhd3eb1b0_0
-streamlit=1.9.0=pyhd8ed1ab_0
-streamlit-aggrid=0.2.3.post2=pypi_0
-streamlit-vega-lite=0.1.0=pypi_0
-tabulate=0.8.9=pypi_0
-tangled-up-in-unicode=0.1.0=pypi_0
-tbb=2021.5.0=h525c30c_0
-tblib=1.7.0=pyhd3eb1b0_0
-tenacity=8.0.1=py39hca03da5_0
-tensorboard-plugin-wit=1.8.1=pypi_0
-terminado=0.13.1=py39hca03da5_0
-testpath=0.5.0=pyhd3eb1b0_0
-threadpoolctl=2.2.0=pyh0d69192_0
-tifffile=2021.7.2=pyhd3eb1b0_2
-tk=8.6.12=he1e0b03_0
-tokenizers=0.11.6=pypi_0
-toml=0.10.2=pyhd3eb1b0_0
-toolz=0.11.2=pyhd3eb1b0_0
-torchvision=0.2.2=py_3
-tornado=6.1=py39h1a28f6b_0
-tqdm=4.64.0=py39hca03da5_0
-traitlets=5.1.1=pyhd3eb1b0_0
-transformers=4.18.0=py39hca03da5_0
-typing-extensions=4.1.1=hd3eb1b0_0
-typing_extensions=4.1.1=pyh06a4308_0
-tzdata=2022a=hda174b7_0
-tzlocal=2.1=py39hca03da5_0
-umap-learn=0.5.3=py39h2804cbe_0
-urllib3=1.26.9=py39hca03da5_0
-validators=0.18.2=pyhd3eb1b0_0
-visions=0.7.4=pypi_0
-watchdog=2.1.6=py39h1a28f6b_0
-wcwidth=0.2.5=pyhd3eb1b0_0
-webencodings=0.5.1=pypi_0
-wheel=0.37.1=pyhd3eb1b0_0
-widgetsnbextension=3.5.2=py39hca03da5_0
-wordcloud=1.8.1=pypi_0
-xarray=0.20.1=pyhd3eb1b0_1
-xxhash=0.8.0=h1a28f6b_3
-xz=5.2.5=h1a28f6b_1
-yake=0.4.8=pypi_0
-yaml=0.2.5=h1a28f6b_0
-yarg=0.1.9=pypi_0
-yarl=1.6.3=py39h1a28f6b_1
-zeromq=4.3.4=hc377ac9_0
-zfp=0.5.5=hc377ac9_6
-zict=2.0.0=pyhd3eb1b0_0
-zipp=3.8.0=py39hca03da5_0
-zlib=1.2.11=h90dfc92_1014
-zstd=1.5.2=h861e0a7_0
+bokeh==2.4.1
+Jinja2==3.1.2
+PyYAML==6.0
+numpy==1.22.4
+packaging==21.3
+Pillow==9.1.1
+tornado==6.1
+typing_extensions==4.2.0
+MarkupSafe==2.1.1
+pyparsing==3.0.9
+pandas==1.4.2
+streamlit==1.2.0