adaptation_specialization_material/plots/ei_sankey.py

import numpy as np
import pandas as pd

import plotly.graph_objects as go
import plotly.io as pio   
pio.kaleido.scope.mathjax = None

import seaborn as sns

import argparse
from os.path import join as opj

parser = argparse.ArgumentParser()
parser.add_argument("--input")
parser.add_argument("--suffix")
parser.add_argument("--compact", action="store_true", default=False)
args = parser.parse_args()

samples = np.load(opj(args.input, f"ei_samples_{args.suffix}.npz"))
topics = pd.read_csv(opj(args.input, "topics.csv"))
topics["label"] = topics["label"].str.replace("\\&", "&", regex=False)
topics = topics[~topics["label"].str.contains("Junk")]["label"].tolist()

n_topics = samples["counts"].shape[1]

source = []
target = []
value = []
color = []
colors = sns.color_palette("hls", n_topics).as_hex()

print(colors)

total_incoming = samples["counts"].mean(axis=0).sum(axis=1)
total_outcoming = samples["counts"].mean(axis=0).sum(axis=0)

incoming_labels = [f"{topics[i]} ({100*total_incoming[i]:.0f}%)" for i in range(n_topics)]
outcoming_labels = [f"{topics[i]} ({100*total_outcoming[i]:.0f}%)" for i in range(n_topics)]

for i in range(n_topics):
    for j in range(n_topics):
        v = samples["counts"][:, i, j].mean()

        x = samples["counts"].mean(axis=0)[i,:].sum()
        y = samples["counts"].mean(axis=0)[:,j].sum()

        value.append(v)
        source.append(i)
        target.append(j + n_topics)
        color.append("rgba(100,100,100,0.2)" if v >= x*y else "rgba(200,200,200,0.02)")

fig = go.Figure(
    data=[
        go.Sankey(
            node=dict(
                pad=7,
                thickness=20,
                line=dict(color="black", width=0.5),
                label=topics + topics,
                color=colors + colors,
            ),
            link=dict(
                source=source,  # indices correspond to labels, eg A1, A2, A1, B1, ...
                target=target,
                value=value,
                color=color,
            ),
        )
    ]
)

fig.update_layout(font_size=13)
fig.update_layout(
    autosize=False,
    width=650 if args.compact else 800,
    height=600,
)
#fig.show()
fig.write_image(opj(args.input, f"sankey_control_{args.suffix}{'_compact' if args.compact else ''}.pdf"), width=650 if args.compact else 800, height=600)

transfers = pd.DataFrame([
    {
        "from": topics[i],
        "to": topics[j],
        "magnitude": 100*samples["counts"].mean(axis=0)[i,j],
        "ratio": samples["counts"].mean(axis=0)[i,j]/samples["counts"].mean(axis=0)[i,:].sum()
    }
    for i in range(n_topics)
    for j in range(n_topics)
])


latex = transfers[transfers["from"]!=transfers["to"]].sort_values("magnitude", ascending=False).head(10).to_latex(
    columns=["from", "to", "magnitude"],
    header=["Origin research area", "Target research area", "Magnitude"],
    index=False,    
    multirow=True,
    multicolumn=True,
    column_format='b{0.4\\textwidth}|b{0.4\\textwidth}|c',
    escape=False,
    float_format=lambda x: f"{x:.2f}",
    caption="Largest transfers across research areas.",
    label="table:largest_transfers"
)

latex = latex.replace('\\\\\n', '\\\\ \\hline\n')

with open(opj(args.input, "largest_transfers.tex"), "w+") as fp:
    fp.write(latex)

latex = transfers[transfers["from"]==transfers["to"]].sort_values("ratio", ascending=False).head(10).to_latex(
    columns=["from", "ratio"],
    header=["Research area", "Conservatism"],
    index=False,    
    multirow=True,
    multicolumn=True,
    column_format='b{0.4\\textwidth}|b{0.4\\textwidth}|c',
    escape=False,
    float_format=lambda x: f"{x:.2f}",
    caption="Most conservative research areas.",
    label="table:most_conservative"
)

latex = latex.replace('\\\\\n', '\\\\ \\hline\n')

with open(opj(args.input, "most_conservative.tex"), "w+") as fp:
    fp.write(latex)