From 3c816ae8dad1b29dfcd9fbcca074eef5278b09cf Mon Sep 17 00:00:00 2001
From: Le Roux Erwan <erwan.le-roux@irstea.fr>
Date: Tue, 21 May 2019 17:56:21 +0200
Subject: [PATCH] [SCM][HYPERCUBE] Add spatial trend visualization with
hypercube
---
.../meteo_france_SCM_study/abstract_study.py | 22 ++++++--
.../hypercube_visualizer.py | 50 +++++++++++++++----
.../main_studies_visualizer.py | 38 ++++++++++----
.../study_visualization/study_visualizer.py | 5 +-
.../abstract_trend_test.py | 10 ++++
.../abstract_margin_function.py | 2 +-
.../margin_fits/plot/create_shifted_cmap.py | 12 +++--
7 files changed, 107 insertions(+), 32 deletions(-)
diff --git a/experiment/meteo_france_SCM_study/abstract_study.py b/experiment/meteo_france_SCM_study/abstract_study.py
index 42973271..25dfd43c 100644
--- a/experiment/meteo_france_SCM_study/abstract_study.py
+++ b/experiment/meteo_france_SCM_study/abstract_study.py
@@ -11,6 +11,8 @@ import numpy as np
import pandas as pd
from PIL import Image
from PIL import ImageDraw
+from matplotlib import cm
+from matplotlib.colors import Normalize
from netCDF4 import Dataset
from experiment.meteo_france_SCM_study.abstract_variable import AbstractVariable
@@ -18,7 +20,7 @@ from experiment.meteo_france_SCM_study.scm_constants import ALTITUDES, ZS_INT_23
from experiment.meteo_france_SCM_study.visualization.utils import get_km_formatter
from extreme_estimator.extreme_models.margin_model.margin_function.abstract_margin_function import \
AbstractMarginFunction
-from extreme_estimator.margin_fits.plot.create_shifted_cmap import get_color_rbga_shifted
+from extreme_estimator.margin_fits.plot.create_shifted_cmap import get_color_rbga_shifted, create_colorbase_axis
from spatio_temporal_dataset.coordinates.abstract_coordinates import AbstractCoordinates
from spatio_temporal_dataset.coordinates.spatial_coordinates.abstract_spatial_coordinates import \
AbstractSpatialCoordinates
@@ -235,12 +237,18 @@ class AbstractStudy(object):
""" Visualization methods """
def visualize_study(self, ax=None, massif_name_to_value=None, show=True, fill=True, replace_blue_by_white=True,
- label=None, add_text=False):
+ label=None, add_text=False, cmap=None):
if massif_name_to_value is None:
massif_name_to_fill_kwargs = None
else:
massif_names, values = list(zip(*massif_name_to_value.items()))
- colors = get_color_rbga_shifted(ax, replace_blue_by_white, values, label=label)
+ if cmap is None:
+ colors = get_color_rbga_shifted(ax, replace_blue_by_white, values, label=label)
+ else:
+ norm = Normalize(0, 100)
+ create_colorbase_axis(ax, label, cmap, norm)
+ m = cm.ScalarMappable(norm=norm, cmap=cmap)
+ colors = [m.to_rgba(value) for value in values]
massif_name_to_fill_kwargs = {massif_name: {'color': color} for massif_name, color in
zip(massif_names, colors)}
@@ -258,8 +266,12 @@ class AbstractStudy(object):
# Potentially, fill the inside of the polygon with some color
if fill and coordinate_id in self.coordinate_id_to_massif_name:
massif_name = self.coordinate_id_to_massif_name[coordinate_id]
- fill_kwargs = massif_name_to_fill_kwargs[massif_name] if massif_name_to_fill_kwargs is not None else {}
- ax.fill(*coords_list, **fill_kwargs)
+ if massif_name_to_fill_kwargs is not None and massif_name in massif_name_to_fill_kwargs:
+ fill_kwargs = massif_name_to_fill_kwargs[massif_name]
+ ax.fill(*coords_list, **fill_kwargs)
+ # else:
+ # fill_kwargs = {}
+
# x , y = list(self.massifs_coordinates.df_all_coordinates.loc[massif_name])
# x , y= coords_list[0][0], coords_list[0][1]
# print(x, y)
diff --git a/experiment/meteo_france_SCM_study/visualization/studies_visualization/hypercube_visualizer.py b/experiment/meteo_france_SCM_study/visualization/studies_visualization/hypercube_visualizer.py
index 99d5c518..e8af0e1a 100644
--- a/experiment/meteo_france_SCM_study/visualization/studies_visualization/hypercube_visualizer.py
+++ b/experiment/meteo_france_SCM_study/visualization/studies_visualization/hypercube_visualizer.py
@@ -6,7 +6,7 @@ import matplotlib.pyplot as plt
import pandas as pd
from experiment.meteo_france_SCM_study.visualization.study_visualization.study_visualizer import StudyVisualizer
-from utils import cached_property, VERSION_TIME
+from utils import cached_property, VERSION_TIME, get_display_name_from_object_type
class HypercubeVisualizer(object):
@@ -17,18 +17,19 @@ class HypercubeVisualizer(object):
"""
def __init__(self, tuple_to_study_visualizer: Dict[Tuple, StudyVisualizer],
- trend_class,
+ trend_test_class,
fast=False,
save_to_file=False):
- self.nb_data_for_fast_mode = 2 if fast else None
+ self.nb_data_for_fast_mode = 7 if fast else None
self.save_to_file = save_to_file
- self.trend_class = trend_class
+ self.trend_test_class = trend_test_class
self.tuple_to_study_visualizer = tuple_to_study_visualizer # type: Dict[Tuple, StudyVisualizer]
# Main attributes defining the hypercube
- def tuple_to_massif_names(self, tuple):
- return self.tuple_to_study_visualizer[tuple].study.study_massif_names
+ @property
+ def trend_test_name(self):
+ return get_display_name_from_object_type(self.trend_test_class)
@cached_property
def starting_years(self):
@@ -40,13 +41,13 @@ class HypercubeVisualizer(object):
@cached_property
def tuple_to_df_trend_type(self):
df_spatio_temporal_trend_types = [
- study_visualizer.df_trend_spatio_temporal(self.trend_class, self.starting_years,
+ study_visualizer.df_trend_spatio_temporal(self.trend_test_class, self.starting_years,
self.nb_data_for_fast_mode)
for study_visualizer in self.tuple_to_study_visualizer.values()]
return dict(zip(self.tuple_to_study_visualizer.keys(), df_spatio_temporal_trend_types))
@cached_property
- def df_hypercube(self):
+ def df_hypercube(self) -> pd.DataFrame:
keys = list(self.tuple_to_df_trend_type.keys())
values = list(self.tuple_to_df_trend_type.values())
df = pd.concat(values, keys=keys, axis=0)
@@ -88,12 +89,12 @@ class AltitudeHypercubeVisualizer(HypercubeVisualizer):
def altitudes(self):
return list(self.tuple_to_study_visualizer.keys())
- def visualize_trend_test(self, ax=None, marker='o'):
+ def visualize_altitude_trend_test(self, ax=None, marker='o'):
if ax is None:
fig, ax = plt.subplots(1, 1, figsize=self.study_visualizer.figsize)
# Plot weighted percentages over the years
- for trend_type, style in self.trend_class.trend_type_to_style().items():
+ for trend_type, style in self.trend_test_class.trend_type_to_style().items():
altitude_percentages = (self.df_hypercube == trend_type)
# Take the mean with respect to the years
altitude_percentages = altitude_percentages.mean(axis=1)
@@ -116,10 +117,37 @@ class AltitudeHypercubeVisualizer(HypercubeVisualizer):
ax.legend()
variable_name = self.study.variable_class.NAME
- title = 'Evolution of {} trends (significative or not) wrt to the altitude'.format(variable_name)
+ name = get_display_name_from_object_type(self.trend_test_class)
+ title = 'Evolution of {} trends (significative or not) wrt to the altitude with {}'.format(variable_name,name)
ax.set_title(title)
self.show_or_save_to_file(specific_title=title)
+ def visualize_spatial_trend_test(self, axes=None):
+ if axes is None:
+ nb_trend_type = len(self.trend_test_class.trend_type_to_style())
+ fig, axes = plt.subplots(1, nb_trend_type, figsize=self.study_visualizer.figsize)
+
+ # Plot weighted percentages over the years
+ for ax, (trend_type, style) in zip(axes, self.trend_test_class.trend_type_to_style().items()):
+ spatial_percentages = (self.df_hypercube == trend_type)
+ # Take the mean with respect to the years
+ spatial_percentages = spatial_percentages.mean(axis=1)
+ # Take the mean with respect the altitude
+ spatial_percentages = spatial_percentages.mean(axis=0, level=1) * 100
+ # Plot values
+ massif_to_value = dict(spatial_percentages)
+ cmap = self.trend_test_class.get_cmap_from_trend_type(trend_type)
+ self.study.visualize_study(ax, massif_to_value, show=False, cmap=cmap, label=None)
+ ax.set_title(trend_type)
+
+ # Global information
+ name = get_display_name_from_object_type(self.trend_test_class)
+ title = 'Repartition of trends (significative or not) with {}'.format(name)
+ title += '\n(in % averaged on altitudes & averaged on starting years)'
+ StudyVisualizer.clean_axes_write_title_on_the_left(axes, title, left_border=None)
+ plt.suptitle(title)
+ self.show_or_save_to_file(specific_title=title)
+
class QuantitityAltitudeHypercubeVisualizer(HypercubeVisualizer):
pass
diff --git a/experiment/meteo_france_SCM_study/visualization/studies_visualization/main_studies_visualizer.py b/experiment/meteo_france_SCM_study/visualization/studies_visualization/main_studies_visualizer.py
index ad30484b..62186a2b 100644
--- a/experiment/meteo_france_SCM_study/visualization/studies_visualization/main_studies_visualizer.py
+++ b/experiment/meteo_france_SCM_study/visualization/studies_visualization/main_studies_visualizer.py
@@ -59,28 +59,44 @@ def altitude_trends_significant():
visualizer.trend_tests_percentage_evolution_with_altitude(trend_test_classes, starting_year_to_weights=None)
-def altitude_hypercube_test():
- save_to_file = False
+def altitude_trend_with_hypercube():
+ save_to_file = True
only_first_one = False
fast = False
altitudes = ALL_ALTITUDES[3:-6]
+ # altitudes = ALL_ALTITUDES[2:4]
+ for study_class in SCM_STUDIES[:]:
+ for trend_test_class in [MannKendallTrendTest, GevLocationTrendTest, GevScaleTrendTest, GevShapeTrendTest][:]:
+ visualizers = [StudyVisualizer(study, temporal_non_stationarity=True, verbose=False, multiprocessing=True)
+ for study in study_iterator(study_class=study_class, only_first_one=only_first_one,
+ altitudes=altitudes)]
+ altitude_to_visualizer = OrderedDict(zip(altitudes, visualizers))
+ visualizer = AltitudeHypercubeVisualizer(altitude_to_visualizer, save_to_file=save_to_file,
+ trend_class=trend_test_class, fast=fast)
+ visualizer.visualize_altitude_trend_test()
+
+def spatial_trend_with_hypercube():
+ save_to_file = False
+ only_first_one = False
+ fast = True
+ # altitudes = ALL_ALTITUDES[3:-6]
altitudes = ALL_ALTITUDES[2:4]
for study_class in SCM_STUDIES[:1]:
- trend_test_class = [MannKendallTrendTest, GevLocationTrendTest, GevScaleTrendTest, GevShapeTrendTest][0]
- visualizers = [StudyVisualizer(study, temporal_non_stationarity=True, verbose=False, multiprocessing=True)
- for study in study_iterator(study_class=study_class, only_first_one=only_first_one,
+ for trend_test_class in [MannKendallTrendTest, GevLocationTrendTest, GevScaleTrendTest, GevShapeTrendTest][:1]:
+ visualizers = [StudyVisualizer(study, temporal_non_stationarity=True, verbose=False, multiprocessing=True)
+ for study in study_iterator(study_class=study_class, only_first_one=only_first_one,
altitudes=altitudes)]
- altitude_to_visualizer = OrderedDict(zip(altitudes, visualizers))
- visualizer = AltitudeHypercubeVisualizer(altitude_to_visualizer, save_to_file=save_to_file,
- trend_class=trend_test_class, fast=fast)
- visualizer.visualize_trend_test()
- # print(visualizer.df_hypercube)
+ altitude_to_visualizer = OrderedDict(zip(altitudes, visualizers))
+ visualizer = AltitudeHypercubeVisualizer(altitude_to_visualizer, save_to_file=save_to_file,
+ trend_test_class=trend_test_class, fast=fast)
+ visualizer.visualize_spatial_trend_test()
+
def main_run():
# altitude_trends()
# altitude_trends_significant()
- altitude_hypercube_test()
+ spatial_trend_with_hypercube()
if __name__ == '__main__':
diff --git a/experiment/meteo_france_SCM_study/visualization/study_visualization/study_visualizer.py b/experiment/meteo_france_SCM_study/visualization/study_visualization/study_visualizer.py
index 32b5898c..98f0aa6f 100644
--- a/experiment/meteo_france_SCM_study/visualization/study_visualization/study_visualizer.py
+++ b/experiment/meteo_france_SCM_study/visualization/study_visualization/study_visualizer.py
@@ -634,7 +634,10 @@ class StudyVisualizer(object):
@staticmethod
def clean_axes_write_title_on_the_left(axes, title, left_border=True):
- if left_border:
+ if left_border is None:
+ clean_axes = axes
+ ax0 = axes[0]
+ elif left_border:
ax0, *clean_axes = axes
else:
*clean_axes, ax0 = axes
diff --git a/experiment/trend_analysis/univariate_trend_test/abstract_trend_test.py b/experiment/trend_analysis/univariate_trend_test/abstract_trend_test.py
index c2afeb6e..eec14042 100644
--- a/experiment/trend_analysis/univariate_trend_test/abstract_trend_test.py
+++ b/experiment/trend_analysis/univariate_trend_test/abstract_trend_test.py
@@ -1,4 +1,5 @@
import random
+import matplotlib.pyplot as plt
from collections import OrderedDict
import numpy as np
@@ -28,6 +29,15 @@ class AbstractTrendTest(object):
d[cls.NO_TREND] = 'k--'
return d
+ @classmethod
+ def get_cmap_from_trend_type(cls, trend_type):
+ if 'positive' in trend_type:
+ return plt.cm.Greens
+ elif 'negative' in trend_type:
+ return plt.cm.Reds
+ else:
+ return plt.cm.binary
+
def __init__(self, years_after_change_point, maxima_after_change_point):
self.years_after_change_point = years_after_change_point
self.maxima_after_change_point = maxima_after_change_point
diff --git a/extreme_estimator/extreme_models/margin_model/margin_function/abstract_margin_function.py b/extreme_estimator/extreme_models/margin_model/margin_function/abstract_margin_function.py
index a2b5b3d0..34cc9b0d 100644
--- a/extreme_estimator/extreme_models/margin_model/margin_function/abstract_margin_function.py
+++ b/extreme_estimator/extreme_models/margin_model/margin_function/abstract_margin_function.py
@@ -7,7 +7,7 @@ import pandas as pd
from experiment.meteo_france_SCM_study.visualization.utils import create_adjusted_axes
from extreme_estimator.margin_fits.gev.gev_params import GevParams
-from extreme_estimator.margin_fits.plot.create_shifted_cmap import plot_extreme_param, imshow_shifted
+from extreme_estimator.margin_fits.plot.create_shifted_cmap import imshow_shifted
from spatio_temporal_dataset.coordinates.abstract_coordinates import AbstractCoordinates
from spatio_temporal_dataset.slicer.split import Split
from utils import cached_property
diff --git a/extreme_estimator/margin_fits/plot/create_shifted_cmap.py b/extreme_estimator/margin_fits/plot/create_shifted_cmap.py
index 194466ec..b605082f 100644
--- a/extreme_estimator/margin_fits/plot/create_shifted_cmap.py
+++ b/extreme_estimator/margin_fits/plot/create_shifted_cmap.py
@@ -23,11 +23,17 @@ def plot_extreme_param(ax, label: str, values: np.ndarray):
cmap = [plt.cm.coolwarm, plt.cm.bwr, plt.cm.seismic][1]
shifted_cmap = shiftedColorMap(cmap, midpoint=midpoint, name='shifted')
norm = mpl.colors.Normalize(vmin=vmin, vmax=vmax)
+ norm = create_colorbase_axis(ax, label, shifted_cmap, norm)
+ return norm, shifted_cmap
+
+
+def create_colorbase_axis(ax, label, cmap, norm):
divider = make_axes_locatable(ax)
cax = divider.append_axes('right', size='5%', pad=0.03)
- cb = cbar.ColorbarBase(cax, cmap=shifted_cmap, norm=norm)
- cb.set_label(label)
- return norm, shifted_cmap
+ cb = cbar.ColorbarBase(cax, cmap=cmap, norm=norm)
+ if isinstance(label, str):
+ cb.set_label(label)
+ return norm
def get_color_rbga_shifted(ax, replace_blue_by_white: bool, values: np.ndarray, label=None):
--
GitLab