# pylint: disable=C0103, too-few-public-methods, locally-disabled, no-self-use, unused-argument, consider-using-enumerate

'''quick plots to view data while debugging'''

from os import path as _path

from warnings import warn as _warn

#import matplotlib as _mpl

import matplotlib.pyplot as Plot

import matplotlib.mlab as _mlab

import numpy as _np

import funclib.baselib as _baselib

from funclib.to_precision import std_notation as _std_notation

from plotlib.mplfuncs import FigWidthsInch as sizes

#import scipy.stats as _stats

#import math as _math

SIZE = (sizes.two_col.value, sizes.two_col.value)

# see C:\Users\Graham Monkman\OneDrive\Documents\PHD\My Papers\Fiducial_error_minimisation\seaborn\mbe_bars.ipynb

# for an example of multifacetted journal formatted histo

def histo(data, bins='auto', normed=True, show=True, figsize=None):

'''(list|ndarray, str|int, bool

Plot a histogram

'''

D = _np.array(data).flatten()

global SIZE

if figsize:

SIZE = figsize

Plot.figure(figsize=SIZE)

dummy, bins, dummy1 = Plot.hist(D, bins=bins, normed=normed)

if normed:

mu = D.mean()

sigma = D.std()

y = _mlab.normpdf(bins, mu, sigma)

dummy = Plot.plot(bins, y, 'r--', linewidth=1)

if show:

Plot.show()

def scatter(x_data, y_data, data_labels=(), group_labels=(), ptsizes=4, data_label_font_sz=8, xlim=None, ylim=None, show=True, figsize=None):

'''(list|tuple|ndarray, list|tuple|ndarray|None,

list|tuple|ndarray, list|tuple|ndarray,

2-tuple|None, 2-tuple|None, bool) - void

Simple x-y scatter plot, supports groups by passing

multiple iterables to x_data and y_data.

x_data:

an iterable of iterable data, or an iterable

eg. [[1,2,4], [10,11,12]]

y_data:

an iterable of iterable data, or an iterable

eg. [[1,2,4], [10,11,12]]

data_labels:

label for each data point

eg. [['a','b','c'], ['x','y','z']]

group_labels:

tuple of group labels to use to create a legend,

midpoint of series labelled.

eg. ['group1', 'group2']

ptsizes:

single value or list-like of point sizes. If

list like, ptsizes matches group_labels by

index. Sensible values are between 1 and 20.

data_label_font_sz:

size of data labels, ie the labels of each point.

8 is usually about right.

xlim:

2-Tuple, lower and upper limits of x, or None

ylim:

2-Tuple, lower and upper limits of y, or None

show:

show the plot

figsize:

2-tuple (w,h) in inches, defaults to qplot.SIZE if None, also set global value

Example:

>>>x_data=[[-1,-2,-3],[3,4,5]]

>>>y_data=[[-5,-6,-7],[5,5,12]]

>>>group_labels=('negatives','positives')

>>>scatter(x_data, y_data, group_labels)

'''

if isinstance(x_data, (list, tuple)):

if _baselib.depth(x_data) == 1:

x_data = [x_data]

if isinstance(y_data, (list, tuple)):

if _baselib.depth(y_data) == 1:

y_data = [y_data]

if isinstance(data_labels, (list, tuple)):

if _baselib.depth(data_labels) == 1:

data_labels = [data_labels]

if not _baselib.isIterable(x_data):

x_data = [_np.array(x_data)]

if not _baselib.isIterable(y_data):

y_data = [_np.array(y_data)]

if not _baselib.isIterable(data_labels):

data_labels = [_np.array(data_labels)]

for ind, x in enumerate(x_data):

if data_labels:

if len(x) != len(data_labels[ind]):

raise ValueError('x_data and data_label lengths must match')

if len(x) != len(y_data[ind]):

raise ValueError('x_data and y_data lengths must match')

if group_labels and len(group_labels) != len(x_data):

print('\nGroup labels: len(group_labels) != len(data). Creating custom labels.')

group_labels = []

_ = [group_labels.append(str(x)) for x in range(len(x_data))]

for ind, x in enumerate(x_data):

if group_labels:

if len(group_labels) >= ind:

grp = '{0!s}'.format(group_labels[ind])

else:

grp = 'grp {0!s}'.format(ind)

if isinstance(x, _np.ndarray):

ndX = x.flatten()

else:

ndX = _np.asarray(x).flatten()

if isinstance(y_data[ind], _np.ndarray):

ndY = y_data[ind].flatten()

else:

ndY = _np.asarray(y_data[ind]).flatten()

global SIZE

if figsize:

SIZE = figsize

Plot.figure(figsize=SIZE)

Plot.scatter(ndX, ndY, s=ptsizes)

axes = Plot.gca()

if isinstance(xlim, (tuple, list)):

if len(xlim) == 2:

axes.set_xlim(xlim)

if isinstance(ylim, (tuple, list)):

if len(ylim) == 2:

axes.set_ylim(ylim)

lst_mean = lambda lst: sum(lst)/len(lst) if lst else 0

#this labels the mid point

if group_labels:

label_pts = [(lst_mean(x), lst_mean(y)) for x, y in zip(x_data, y_data)]

for i in range(len(label_pts)):

Plot.annotate(group_labels[i], xy=label_pts[i], xytext=label_pts[i], horizontalalignment='center', verticalalignment='center', size=data_label_font_sz)

if data_labels:

for n, labels in enumerate(data_labels):

for i, label in enumerate(labels):

lbl_pt = (x_data[n][i], y_data[n][i])

Plot.annotate(label, xy=lbl_pt, xytext=lbl_pt, horizontalalignment='center', verticalalignment='center', size=data_label_font_sz)

#Plot.legend(groups, loc='upper right')

if show:

Plot.show()

def bar_(xvalues, yvalues, title='', xlabel='x', ylabel='y', alpha=1, color='royalblue', xlabels=None, width=0.8, output=None, xmax=None, ymax=None, vlines=None, show=True):

'''

Create and save a bar plot.

Args:

- xvalues = x-axis positions for bars

- yvalues = y-axis magnitudes of each bar

- title = title of plot. Also used for filename

- xlabel = x-axis label

- ylabel = y-axis label

Options:

- alpha = opacity of bars

- color = color of bars

- xlabels = x-axis labels for each bar

- xmax = max x-value

- ymax = max y-value

- width = width of bars

'''

_, ax1 = Plot.subplots()

# Size

if xmax:

xmin, _ = ax1.get_xlim()

ax1.set_xlim(xmin, xmax)

if ymax:

ymin, _ = ax1.get_ylim()

ax1.set_ylim(ymin, ymax)

# Add data

dummy = Plot.bar(xvalues, yvalues, width=width, color=color, alpha=alpha)

# Add extra lines

for line in (vlines or []):

Plot.axvline(line["xpos"], color=line["color"],

linestyle=line["style"], linewidth=line["width"])

# Labels

if xlabels:

Plot.xticks([x + width / 2 for x in xvalues], xlabels)

ax1.set_title(title)

ax1.set_xlabel(xlabel)

ax1.set_ylabel(ylabel)

# Save

if output:

try:

Plot.savefig(_path.normpath(output))

print("Saved bar chart to '%s'" % output)

except Exception as _:

_warn('Faile to save graph to %s. Check the folder is valid.' % output)

if show:

Plot.show()

Plot.clf()

Plot.close()

def pretty_bin(bin_edges, use_mid=True, precision=2):

'''(list|tuple, bool) -> list

Given bin edges, get text x col labels

bin_edges: a list/tuple with the bin edges, eg.

[0.0, 5, 1.0]

use_mid: Generate midpoint labels, rather than ranges

[0.25, 0.75]

Example:

>>>pretty_bin([0, 0.5, 1])

[0.25, 0.75]

>>>pretty_bin([0, 0.5, 1], use_mid=False)

[0.0-0.50, 0.5-1.0]

'''

out = []

for i, item in enumerate(bin_edges[0:len(bin_edges)-1]):

if use_mid:

v = ((bin_edges[i+1] - item) * 0.5) + item

v = _std_notation(v, precision)

out.append(str(v))

else:

a = _std_notation(item, precision)

b = _std_notation(bin_edges[i+1], precision)

out.append('%s-%s' % (a, b))

return out