# Source code for pytesmo.time_series.plotting

# Copyright (c) 2014,Vienna University of Technology,
# Department of Geodesy and Geoinformation

# Redistribution and use in source and binary forms, with or without
# modification, are permitted provided that the following conditions are met:
#   * Redistributions of source code must retain the above copyright
#     notice, this list of conditions and the following disclaimer.
#    * Redistributions in binary form must reproduce the above copyright
#      notice, this list of conditions and the following disclaimer in the
#      documentation and/or other materials provided with the distribution.
#    * Neither the name of the Vienna University of Technology,
#      Department of Geodesy and Geoinformation nor the
#      names of its contributors may be used to endorse or promote products
#      derived from this software without specific prior written permission.

# THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS" AND
# ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED
# WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE
# DISCLAIMED. IN NO EVENT SHALL VIENNA UNIVERSITY OF TECHNOLOGY,
# DEPARTMENT OF GEODESY AND GEOINFORMATION BE LIABLE FOR ANY
# DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES
# (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES;
# LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND
# ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
# (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS
# SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.

'''
Created on Mar 7, 2014

Plot anomalies around climatology using colors

@author: Christoph Paulik christoph.paulik@geo.tuwien.ac.at
'''

import matplotlib.pyplot as plt
import matplotlib.gridspec as gridspec
import pandas as pd

import pytesmo.time_series.anomaly as anom

[docs]def plot_clim_anom(df, clim=None, axes=None, markersize=0.75,
mfc='0.3', mec='0.3', clim_color='0.0',
clim_linewidth=0.5, clim_linestyle='-',
"""
Takes a pandas DataFrame and calculates the climatology and anomaly
and plots them in a nice way for each column

Parameters
----------
df : pandas.DataFrame
clim : pandas.DataFrame, optional
if given these climatologies will be used
if not given then climatologies will be calculated
this DataFrame must have the same number of columns as df
and also the column names.
each climatology must have doy as index.
axes : list of matplotlib.Axes, optional
list of axes on which each column should be plotted
if not given a standard layout is generated
markersize : float, optional
size of the markers for the datapoints
mfc : matplotlib color, optional
markerfacecolor, color of the marker face
mec : matplotlib color, optional
markeredgecolor
clim_color : matplotlib color, optional
color of the climatology
clim_linewidth : float, optional
linewidth of the climatology
clim_linestyle : string, optional
linestyle of the climatology
pos_anom_color : matplotlib color, optional
color of the positive anomaly
neg_anom_color : matplotlib color, optional
color of the negative anomaly
anom_linewidth : float, optional
linewidth of the anomaly lines
if set each subplot will have it's column name as title
Default : True

Returns
-------
Figure : matplotlib.Figure
if no axes were given
axes : list of matploblib.Axes
if no axes were given
"""

if type(df) == pd.Series:
df = pd.DataFrame(df)

nr_columns = len(df.columns)

# make own axis if necessary
if axes is None:
own_axis = True
gs = gridspec.GridSpec(nr_columns, 1, right=0.8)

fig = plt.figure(num=None, figsize=(6, 2 * nr_columns),
dpi=150, facecolor='w', edgecolor='k')

axes = []
for i, grid in enumerate(gs):
if i < nr_columns - 1:
axes.append(ax)
ax.xaxis.set_visible(False)
axes.append(last_axis)

else:
own_axis = False

for i, column in enumerate(df):
Ser = df[column]
ax = axes[i]

if clim is None:
clima = anom.calc_climatology(Ser)
else:
clima = pd.Series(clim[column])
anomaly = anom.calc_anomaly(Ser, climatology=clima, return_clim=True)

anomaly[Ser.name] = Ser
anomaly = anomaly.dropna()

pos_anom = anomaly[Ser.name].values > anomaly['climatology'].values
neg_anom = anomaly[Ser.name].values < anomaly['climatology'].values

ax.plot(anomaly.index, anomaly[Ser.name].values, 'o',
markersize=markersize, mfc=mfc, mec=mec)

ax.plot(anomaly.index, anomaly['climatology'].values,
linestyle=clim_linestyle,
color=clim_color,
linewidth=clim_linewidth)

ax.fill_between(anomaly.index,
anomaly[Ser.name].values,
anomaly['climatology'].values, interpolate=True,
where=pos_anom, color=pos_anom_color,
linewidth=anom_linewidth)
ax.fill_between(anomaly.index,
anomaly[Ser.name].values,
anomaly['climatology'].values, interpolate=True,
where=neg_anom, color=neg_anom_color,
linewidth=anom_linewidth)