2.1 viz h
import sys, os, pathlib
from pyprojroot import here
root = here(project_files=[".here"])
sys.path.append(str(here()))
PROJ_PATH = pathlib.Path(root)
import scipy.io as scio
# NUMPY SETTINGS
import numpy as np
np.set_printoptions(precision=3, suppress=True)
# MATPLOTLIB Settings
import matplotlib as mpl
import matplotlib.pyplot as plt
%matplotlib inline
%config InlineBackend.figure_format = 'retina'
# SEABORN SETTINGS
import seaborn as sns
sns.set_context(context='talk',font_scale=0.7)
# sns.set(rc={'figure.figsize': (12, 9.)})
# sns.set_style("whitegrid")
# PANDAS SETTINGS
import pandas as pd
pd.set_option("display.max_rows", 120)
pd.set_option("display.max_columns", 120)
# LOGGING SETTINGS
import sys
import logging
logging.basicConfig(
level=logging.INFO,
stream=sys.stdout,
format='%(asctime) s:%(levelname) s:%(message)s '
)
logger = logging.getLogger()
#logger.setLevel(logging.INFO)
%load_ext autoreload
%autoreload 2
DATA_PATH = PROJ_PATH.joinpath("data/spa_temp/info_earth/entropy")
# DATA_NEW_PATH = PROJ_PATH.joinpath("data/spa_temp/entropy/old_results")
FIG_PATH = PROJ_PATH.joinpath("docs/pics/entropy/")
file_name = '2019_CI_FIG_ENTR_RBIG_2001_'
variables = {
'gpp': 'gross_primary_productivity',
'lai': 'leaf_area_index',
'lst': 'land_surface_temperature',
'precip': 'precipitation',
'rm': 'root_moisture',
'sm': 'soil_moisture',
}
rs_dims = [
1, 4, 9, 16
]
dims = [
1, 4, 9, 16, 25, 36, 49
]
!ls $DATA_PATH
def plot_entropy(results_df, variable, rs=False, save_name=None):
fig, ax = plt.subplots(figsize=(7,5))
# sns.lineplot(ax=ax, x="I", y="n_features", data=fmri)
sub_df = results_df[results_df['variable'] == variable]
# print(sub_df)
sns.lineplot(
ax=ax, x='ratio', y='entropy',
data=sub_df, hue=sub_df.n_dimensions.values,
marker='o',markersize=10)
# sns.lineplot(ax=ax, x='nm_features', y='I', data=ndvi_df, label=label_main, linewidth=4,)
ax.set_xlabel('Ratio', )
ax.set_ylabel('Entropy')
plt.legend(rs_dims if rs is True else dims)
plt.tight_layout()
if save_name is not None:
fig.savefig(FIG_PATH.joinpath(f"{save_name}.png"), frameon=False, )
else:
plt.show()
Results I - Resampled, Original¶
results_df = pd.DataFrame()
version = 'v1'
region = 'world'
period = '2002_2010'
n_samples = 200
rs = True
rc = False
for ivariable in variables.keys():
save_name = f"{version}_{region}_{ivariable}_{period}_s{n_samples}k"
if rs:
save_name += "_rs1MS"
if rc:
save_name += "_rc"
print(ivariable)
results_df = pd.read_csv(DATA_PATH.joinpath(save_name + ".csv"), index_col=0)
results_df['ratio'] = (results_df['spatial']**2 / results_df['temporal']) / results_df['n_dimensions']
results_df['entropy'] /= results_df['n_dimensions']
plot_entropy(results_df, variables[ivariable], rs, save_name)
Results II - Resampled, Climatology Removed¶
results_df = pd.DataFrame()
version = 'v1'
region = 'world'
period = '2002_2010'
n_samples = 200
rs = True
rc = True
for ivariable in variables.keys():
save_name = f"{version}_{region}_{ivariable}_{period}_s{n_samples}k"
if rs:
save_name += "_rs1MS"
if rc:
save_name += "_rc"
print(ivariable)
results_df = pd.read_csv(DATA_PATH.joinpath(save_name + ".csv"), index_col=0)
results_df['ratio'] = (results_df['spatial']**2 / results_df['temporal']) / results_df['n_dimensions']
results_df['entropy'] /= results_df['n_dimensions']
plot_entropy(results_df, variables[ivariable], rs, save_name)
# for idims, (ratio, entropy) in zip(dims, data_mat['RES'][0]):
# for iratio, ientropy in zip(np.nditer(ratio), np.nditer(entropy)):
# # print(type(idims))
# # print(type(iratio))
# # print(type(ientropy))
# results_df = results_df.append({
# 'variable': ivariable,
# 'ratio': float(iratio),
# 'h': float(ientropy),
# 'dims': float(idims)
# }, ignore_index=True)
# results_df.to_csv(DATA_NEW_PATH.joinpath("old_results.csv"))
# len(precip_mat['RES'][0])
Results III - Full¶
results_df = pd.DataFrame()
version = 'v2'
region = 'world'
period = '2002_2010'
n_samples = 200
rs = False
rc = False
for ivariable in variables.keys():
save_name = f"{version}_{region}_{ivariable}_{period}_s{n_samples}k"
if rs:
save_name += "_rs1MS"
if rc:
save_name += "_rc"
print(ivariable)
results_df = pd.read_csv(DATA_PATH.joinpath(save_name + ".csv"), index_col=0)
results_df['ratio'] = (results_df['spatial']**2 / results_df['temporal']) / results_df['n_dimensions']
results_df['entropy'] /= results_df['n_dimensions']
plot_entropy(results_df, variables[ivariable], rs, save_name)
Results IV - Full, Climatology Removed¶
results_df = pd.DataFrame()
version = 'v2'
region = 'world'
period = '2002_2010'
n_samples = 200
rs = False
rc = True
for ivariable in variables.keys():
save_name = f"{version}_{region}_{ivariable}_{period}_s{n_samples}k"
if rs:
save_name += "_rs1MS"
if rc:
save_name += "_rc"
print(ivariable)
results_df = pd.read_csv(DATA_PATH.joinpath(save_name + ".csv"), index_col=0)
results_df['ratio'] = (results_df['spatial']**2 / results_df['temporal']) / results_df['n_dimensions']
results_df['entropy'] /= results_df['n_dimensions']
plot_entropy(results_df, variables[ivariable], rs, save_name)
