from Thermal_Models import read_database, clean_paired_data, gmean, normalise, weighted_std, weighted_amean
import numpy as np
import pandas as pd
print('\n\tCalculating EG values...\n')
agg_data = read_database('../Data/summaries/non_aggregate_data_fix.csv')
gb_cols = [['ConKingdom'], ['RespiratoryPathway'], ['TempPref', 'ConKingdom'],
['RespiratoryPathway', 'TempPref']]
#calculate ES as the weighted geometric mean
def ES(g):
pair_data = agg_data.ix[g.index]['E_std']
return weighted_amean(g, pair_data, normalise_weights=True)
#calculate ES certainty as the weighted standard deviation
def ES_min(g):
pair_data = np.array(agg_data.ix[g.index]['E_std'])
g = np.array(g)
length = len(g)
bootvals = []
for i in range(1000):
indices = np.random.choice(length, length)
g_boot = g[indices]
pair_boot = pair_data[indices]
bootvals.append(
estimate_uncertainty, \
bootstrap_model
import numpy as np
import pandas as pd
from datetime import datetime
import warnings
warnings.simplefilter(
action="ignore", category=FutureWarning
) #Probably ought to deal with this, but for the moment I'll sweep it under the rug.
warnings.simplefilter(action="ignore", category=RuntimeWarning)
starttime = datetime.now()
data_path = '../Data/database.csv'
data = read_database(data_path)
Datasets = split_datasets(
data) #return a dictionary of growth curves split by ID
#names of the models I want to fit as strings
model_names = ['schoolfield_two_factor_tpk']
all_models = [] #place to put fitted models
#Column names of parameters I want to use as explanatory variables when I analyse the data
aux_parameters = [
'FinalID', 'OriginalID', 'Citation', 'Latitude', 'Longitude', 'ConKingdom',
'ConPhylum', 'ConClass', 'ConOrder', 'ConFamily', 'ConGenus', 'ConSpecies',
'OptimalConditions', 'ConLabGrowthTemp'
]
for i in Datasets.keys(): #Iterate through the dictionary by key
import pandas as pd
import numpy as np
from Thermal_Models import read_database, clean_paired_data, gmean, normalise, weighted_std, weighted_amean
print('\n\tAggregating Summary...\n')
def mesophile_thermophile(entry):
if entry > 323.15:
return 'Thermophile'
else:
return 'Mesophile'
agg_data = read_database('../Data/summaries/summary.csv')
agg_data = agg_data[(agg_data.Citation != "Mackey et al. (2013)")] # I really don't trust this data having looked at it (it came from BioTraits, not us) - TS
colnames = list(agg_data.columns.values)
#columns to keep
keep = ['Species', 'Trait', 'ConKingdom', 'ConPhylum', 'ConClass', 'ConOrder', 'ConFamily', 'ConGenus', 'RespiratoryPathway', 'Max_response', 'Est_Tpk', 'Rank', 'Points_Before_Peak', 'Tpk_std', 'Response_std', 'E_std', 'E']
#Drop any column not in the keep list
for colname in colnames:
if colname not in keep:
agg_data.drop(colname, 1, inplace=True)
#select best fits for growth rates only
groups = agg_data[(agg_data.Rank == 1) & (agg_data.Trait == 'Specific Growth Rate') & (agg_data.Points_Before_Peak > 2)]
#Rename appropriately
groups.rename(columns={'Est_Tpk': 'Est_Tpk',
'Tpk_std': 'Est_Tpk_std',
'Max_response': 'Est_Response',
bootstrap_model, \
split_datasets, \
rank_and_flatten, \
compile_models, \
find_linear_arrhenius, \
estimate_uncertainty
import numpy as np
import pandas as pd
from datetime import datetime
import matplotlib.pyplot as plt
starttime = datetime.now()
summary_path = '../Data/summaries/group_means_non_aggregated.csv' #We will merge results into this
data_path = '../Data/summaries/non_aggregate_data_fix.csv'
data = read_database(data_path)
summary = read_database(summary_path)
data = data[(np.isfinite(data.Est_Tpk)) & (np.isfinite(data.Est_Response))]
aux_parameters = [] #no supplementary information needed
analysis_levels = ['ConKingdom'] #levels at which models should be fitted
model_names = ['boltzmann_arrhenius_one_weight_non_linear'] #models to fit
for level in analysis_levels:
if level == '':
hist_axes = True
else:
hist_axes = False
#see readme for more details on this dictionary
param_est_flags = {
from Thermal_Models import read_database, clean_paired_data, gmean, normalise, weighted_std, weighted_amean
import numpy as np
import pandas as pd
print('\n\tCalculating EG values...\n')
agg_data = read_database(
'../Data/summaries/non_aggregate_data_fluxes_best.csv')
gb_cols = ['ConKingdom']
#calculate ES as the weighted geometric mean
def ES(g):
pair_data = agg_data.ix[g.index]['E_std']
return weighted_amean(g, pair_data, normalise_weights=True)
#calculate ES certainty as the weighted standard deviation
def ES_min(g):
pair_data = np.array(agg_data.ix[g.index]['E_std'])
g = np.array(g)
length = len(g)
bootvals = []
for i in range(1000):
indices = np.random.choice(length, length)
g_boot = g[indices]
pair_boot = pair_data[indices]
bootvals.append(
schoolfield_original, \
LM, \
read_database, \
fit_models, \
bootstrap_model, \
split_datasets, \
rank_and_flatten, \
compile_models
import numpy as np
import pandas as pd
from datetime import datetime
starttime = datetime.now()
data_path = '../Data/summaries/aggregate_data.csv'
data = read_database(data_path)
aux_parameters = [] #no supplementary information needed
analysis_levels = ['ConKingdom', 'ConPhylum', 'ConClass', 'ConOrder', 'ConFamily', 'ConGenus'] #levels at which models should be fitted
model_names = ['Boltzmann_Arrhenius', 'LM'] #models to fit
for level in analysis_levels:
#set the axis type of the plot based on analysis level.
if level == 'ConKingdom':
hist_axes = True
else:
hist_axes = False
#see readme for more details on this dictionary
param_est_flags = {'Trait_Col_Name': 'Trait',
'X_vals_col_name': 'Est.Tpk',