🎨📝✅💚⬆️ Add geom_median to env, move favicon assets

📝🎨 Add Bioconda project url to setup.py
🎨 Add toast when browser storage limits are reached (issue #11)
:art::memo: Move favicon to automappa/assets
:arrow_up::art: Add geom-median to environment.yml
:art: Add get_clusters_geom_medians(...) to celery tasks
:arrow_up::art: Bump VERSION to 2.2.0
:see_no_evil: Add data to .gitignore

.gitignore

@@ -9,4 +9,5 @@ app/__pycache__
 dist
 build
 Automappa.egg-info
-.env
+.env
+data

VERSION

@@ -1 +1 @@
-2.1.0
+2.2.0

automappa/assets/site.webmanifest

@@ -0,0 +1 @@
+{"name":"","short_name":"","icons":[{"src":"/android-chrome-192x192.png","sizes":"192x192","type":"image/png"},{"src":"/android-chrome-512x512.png","sizes":"512x512","type":"image/png"}],"theme_color":"#ffffff","background_color":"#ffffff","display":"standalone"}

automappa/celery.py

@@ -1,12 +1,15 @@
 #!/usr/bin/env python
 
 
+
 import os
 import glob
+import pandas as pd
 
-from autometa.common.external import hmmscan
+from geom_median.numpy import compute_geometric_median
 from celery import Celery, chain
 from dotenv import load_dotenv
+from autometa.common.external import hmmscan
 
 load_dotenv()
 
@@ -21,6 +24,7 @@
 # TODO: Marker symbols
 # TODO: CheckM annotation
 # TODO: kmer freq. analysis pipeline
+# TODO: scatterplot 2-d embedding views
 
 @CELERY.task
 def hmmdb_formatter(hmmdb) -> None:
@@ -69,3 +73,36 @@ def scanner(seqfile, hmmdb, out) -> str:
         out = os.path.join(outdir, outfilename)
         hmmdb_formatter.s(hmmdb).apply_async()
         scanner.s(seqfile, hmmdb, out).apply_async(countdown=2)
+
+
+def get_clusters_geom_medians(df: pd.DataFrame, cluster_col: str = "cluster", weight_col: str='length') -> pd.DataFrame:
+    """Compute each cluster's (`cluster_col`) geometric median weighted by contig length (`weight_col`)
+
+    Parameters
+    ----------
+    df : pd.DataFrame
+        Table containing x_1 and x_2 coordinates and `cluster_col` from embedding
+    cluster_col : str, optional
+        Value to use for cluster column, by default "cluster"
+    weight_col : str, optional
+        Column to use for weighting the geometric median computation, by default 'length'
+
+    Returns
+    -------
+    pd.DataFrame
+        index=range(cluster_1, cluster_n), cols=[cluster_col, x_1, x_2, termination, weighted]
+        `x_1` and `x_2` correspond to the computed geometric median values corresponding to the respective cluster.
+        `termination` is the reason of termination passed from the `compute_geometric_median` function
+        `weighted` denotes the value that was used for weighting the cluster's geometric median (`weight_col`)
+
+    """
+    medians = []
+    for cluster, dff in df.groupby(cluster_col):
+        points = dff[['x_1', 'x_2']].to_numpy()
+        if weight_col:
+            weights = dff[weight_col].to_numpy()
+        out = compute_geometric_median(points=points, weights=weights)
+        median = out.median
+        medians.append({cluster_col: cluster, "x_1": median[0], "x_2": median[1], "termination": out.termination, "weighted":weight_col})
+    medians_df = pd.DataFrame(medians)
+    return medians_df

automappa/index.py

@@ -115,12 +115,25 @@ def main():
 
     # Check dataset size for dcc.Store(...) with browser limits...
     # For details see: https://stackoverflow.com/a/61018107 and https://arty.name/localstorage.html
-    chrome_browser_quota = 5200000
+    chrome_browser_quota = 5_200_000
     dataset_chars = len(binning.to_json(orient="split"))
     if dataset_chars >= chrome_browser_quota:
-        logger.warning(f"{args.binning_main} exceeds browser storage limits ({dataset_chars} > {chrome_browser_quota}).")
+        logger.warning(f"{args.binning_main} exceeds browser storage limits ({dataset_chars:,} > {chrome_browser_quota:,}).")
         logger.warning("Persisting refinements is DISABLED!")
 
+        browser_storage_toast = dbc.Toast(
+            f"{args.binning_main} exceeds browser storage limits ({dataset_chars:,} > {chrome_browser_quota:,}).",
+            id="positioned-toast",
+            header="Persisting refinements DISABLED",
+            is_open=True,
+            dismissable=True,
+            icon="danger",
+            # top: 66 positions the toast below the navbar
+            style={"position": "fixed", "top": 66, "right": 10, "width": 350},
+        )
+    else:
+        browser_storage_toast = dbc.Toast(is_open=False)
+
     # Metagenome Annotations Store
     metagenome_annotations_store = dcc.Store(
         id="metagenome-annotations",
@@ -194,6 +207,7 @@ def main():
             dbc.Tabs(
                 id="tabs", children=[refinement_tab, summary_tab], className="nav-fill"
             ),
+            html.Div(browser_storage_toast),
             html.Div(id="tab-content"),
         ],
         fluid=True,

automappa/utils/figures.py

@@ -6,7 +6,6 @@
 from dash.exceptions import PreventUpdate
 from plotly import graph_objects as go
 
-
 def taxonomy_sankey(df: pd.DataFrame, selected_rank: str = "species") -> go.Figure:
     ranks = ["superkingdom", "phylum", "class", "order", "family", "genus", "species"]
     n_ranks = len(ranks[: ranks.index(selected_rank)])
@@ -113,6 +112,8 @@ def marker_size_scaler(x: pd.DataFrame, scale_by: str = "length") -> int:
     return x_scaled
 
 
+
+
 def get_scatterplot_2d(
     df,
     x_axis: str = "x_1",

environment.yml

@@ -16,4 +16,10 @@ dependencies:
   - numpy
   - pandas
   - plotly
+<<<<<<< Updated upstream
   - python-dotenv
+=======
+  - pip
+  - pip:
+    - geom-median
+>>>>>>> Stashed changes

setup.py

@@ -20,7 +20,7 @@ def read(fname):
     python_requires=">=3.7",
     version=version,
     packages=find_packages(exclude=["tests"]),
-    package_data={"": ["app/assets/*"]},
+    package_data={"": ["assets/*"]},
     entry_points={
         "console_scripts": [
             "automappa = automappa.__main__:main",
@@ -32,6 +32,10 @@ def read(fname):
     long_description=long_description,
     long_description_content_type="text/markdown",
     url="https://github.com/WiscEvan/Automappa",
+    project_urls={
+        "Bug Tracker": "https://github.com/WiscEvan/Automappa/issues",
+        "Bioconda": "https://anaconda.org/bioconda/automappa",
+    },
     license="GNU Affero General Public License v3 or later (AGPLv3+)",
     classifiers=[
         "Programming Language :: Python",