From 84db01b6557fa73fd40682361f42e0c85dfce7b7 Mon Sep 17 00:00:00 2001
From: Peter Swain <peter.swain@ed.ac.uk>
Date: Wed, 7 Aug 2024 10:59:42 +0100
Subject: [PATCH] change(plotting): added plot_binned_mean
---
src/wela/plotting.py | 67 ++++++++++++++++++++++++++++++++++++++++++++
1 file changed, 67 insertions(+)
diff --git a/src/wela/plotting.py b/src/wela/plotting.py
index 64d04c8..5c95d8c 100644
--- a/src/wela/plotting.py
+++ b/src/wela/plotting.py
@@ -1,9 +1,12 @@
+"""Plotting routines to work with dataloader."""
+
from copy import copy
import matplotlib.cm
import matplotlib.pylab as plt
import numpy as np
import numpy.matlib
+from scipy.stats import binned_statistic
try:
from sklearn.preprocessing import StandardScaler
@@ -754,3 +757,67 @@ def get_bud_to_bud_data(
local_signals.append(local_data)
local_times.append(t[start_tpt_i : end_tpt_i + 1])
return local_signals, local_times
+
+
+def plot_binned_mean(df, x_signal, y_signal, bins=10, groups=None, fmt="o-"):
+ """
+ Plot the mean of y_signal found for bins of x_signal against x_signal.
+
+ Use scipy's binned_statistic.
+
+ Parameters
+ ----------
+ df: pd.DataFrame
+ Dataframe with the data, typically dl.df.
+ x_signal: str
+ Name of the signal to bin and plot on the x-axis.
+ y_signal: str
+ Name of the signal to be averaged in bins of x_signal.
+ bins: int
+ Number of bins.
+ groups: list of str (optional)
+ Specific groups to plot.
+ fmt: str (optional)
+ Formatting for points and lines, passed to plt.errorbar.
+
+ Example
+ -------
+ >>> plot_binned_mean(dl.df, "median_GFP", "bud_growth_rate",
+ bins=10, groups=["2pc_raf", "2pc_glc"])
+ """
+ stats_dict = {}
+ if groups is None:
+ groups = df.group.unique()
+ for group in groups:
+ sdf = df[df.group == group][[x_signal, y_signal]].dropna()
+ stats = ["mean", "median", "std", "count"]
+ for stat in stats:
+ stats_dict[f"{stat}_{group}"], bin_edges, _ = binned_statistic(
+ sdf.median_GFP.values,
+ values=sdf.bud_growth_rate.values,
+ statistic=stat,
+ bins=bins,
+ )
+ stats_dict[f"stderr_{group}"] = stats_dict[f"std_{group}"] / np.sqrt(
+ stats_dict[f"count_{group}"]
+ )
+ stats_dict[f"bin_midpoints_{group}"] = np.array(
+ [
+ np.mean([bin_edges[i], bin_edges[i + 1]])
+ for i in range(len(bin_edges) - 1)
+ ]
+ )
+ # plot using errorbar
+ plt.figure()
+ for group in groups:
+ plt.errorbar(
+ stats_dict[f"bin_midpoints_{group}"],
+ stats_dict[f"mean_{group}"],
+ yerr=stats_dict[f"stderr_{group}"],
+ fmt=fmt,
+ label=group,
+ )
+ plt.xlabel(x_signal.replace("_", " "))
+ plt.ylabel(y_signal.replace("_", " "))
+ plt.legend()
+ plt.show(block=False)
--
GitLab