'Name', 'Date', 'Open', 'High', 'Low', 'Close', 'Volume', 'Market Cap'
]
df = {}
df = pd.DataFrame(columns=columns)
col_time = ['DBegin', 'DEnd', 'Type', 'VBegin', 'VEnd', 'Volume', 'Mean']
dl = {}
dl = pd.DataFrame(columns=col_time)
print cryptoconcurrenciesName
for cryptoName in cryptoconcurrenciesName:
file = mypath + cryptoName
cryptoconcurrencies = rp.loadDataCSV(file,
target=rp.NONE,
null_target_procedure=rp.DELETE_ROW,
null_procedure=rp.MEAN,
na_values='-')
df = df.append(cryptoconcurrencies.dataFrame)
df = df.fillna(cryptoName[0:-4])
monedas = ['Bitcoin'] # ,'Ethereum','Monero']
bamboo = Bamboo('Coins', df, columns, target='Market Cap')
bambooList = rp.divideDataFrame(bamboo, 'Name')
for moneda in monedas:
for coso in bambooList:
# print coso.dataFrame['Name'].iloc[0]
if coso.dataFrame['Name'].iloc[0] == moneda:
cdf = coso.dataFrame
#!/usr/bin/env python2
# -*- coding: utf-8 -*-
"""
Created on Wed Nov 15 16:12:35 2017
@author: edu
"""
import RedPandas as RP
weeks, features, n_nulls = RP.loadDataCSV(name='dengue_train.csv',
target='total_cases',
null_target_procedure=RP.DELETE_ROW,
null_procedure=RP.MEAN)
#divide into iq and sj
# set the index to be this and don't drop
#weeks.set_index(keys=['city'], drop=False,inplace=True)
# get a list of city names
cityNames = weeks['city'].unique().tolist()
iqWeeks = weeks.loc[weeks.city == 'iq']
sjWeeks = weeks.loc[weeks.city == 'sj']
#0 Delete outliers
#1st iteration
iqWeeks.drop(iqWeeks.index[[244, 104, 3, 103, 51, 306, 10, 115, 273]],
inplace=True)
sjWeeks.drop(sjWeeks.index[[507, 500, 705, 800]], inplace=True)
#2nd iteration
iqWeeks.drop(iqWeeks.index[[23]], inplace=True)
# -*- coding: utf-8 -*-
# 1. Load data
import RedPandas as RP
data, name, n_nulls = RP.loadDataCSV(name='dengue_train.csv',
target='total_cases',
null_target_procedure=RP.DELETE_ROW,
null_procedure=RP.MEAN)
name = list(data.head(0))
import numpy
#divide into iq and sj
'''
# set the index to be this and don't drop
data.set_index(keys=['city'], drop=False,inplace=True)
'''
# get a list of city names
cityNames = data['city'].unique().tolist()
iqWeeks = data.loc[data.city == 'iq']
sjWeeks = data.loc[data.city == 'sj']
name.remove('total_cases')
name.remove('city')
name.remove('year')
name.remove('week_start_date')
#0 Delete outliers
#1st iteration
iqWeeks.drop(iqWeeks.index[[244, 104, 3, 103, 51, 306, 10, 115, 273]],
inplace=True)
# -*- coding: utf-8 -*-
"""
Created on Wed Dec 06 16:00:23 2017
@author: Edu
"""
import RedPandas as rp
import numpy as np
import pandas as pd
sj_train = rp.loadDataCSV('db/sj_train.csv', target='total_cases')
sj_test = rp.loadDataCSV('db/sj_test.csv', target='total_cases')
#San Juan analysis
sj_train.features.remove('year')
sj_train.features.remove('total_cases')
sj_train.features.remove('city')
sj_train.features.remove('week_start_date')
#CV tests
#Decision tree
#This is mostly for feature selection
min_range = 2
max_range = 30
depths = []
#Multiple CV tests done since it's random so we get the most frequent depth
for x in range(0, 1):
# -*- coding: utf-8 -*-
import RedPandas as rp
den_train = rp.loadDataCSV('db/dengue_train.csv',
target='total_cases',
null_target_procedure=rp.DELETE_ROW,
null_procedure=rp.MEAN)
den_test = rp.loadDataCSV('db/dengue_test.csv',
null_target_procedure=rp.DELETE_ROW,
null_procedure=rp.MEAN)
#Dividing the training into 2 dataframes, Iquitos and San Juan
den_train_div = rp.divideDataFrame(den_train, 'city')
for div in den_train_div:
if div.dataFrame['city'].iloc[0] == 'iq':
iq_train = div
elif div.dataFrame['city'].iloc[0] == 'sj':
sj_train = div
div.reportBasicInfo(printOnScreen=False)
#Dividing the test into 2 dataframes, Iquitos and San Juan
den_test_div = rp.divideDataFrame(den_test, 'city')
#!/usr/bin/env python2
# -*- coding: utf-8 -*-
"""
Created on Tue Nov 21 12:57:30 2017
@author: edu
"""
# 0 load data
import RedPandas as rp
import pandas as pd
import matplotlib.pyplot as plt
weeks, features, n_nulls = rp.loadDataCSV(name='dengue_train.csv',
target='total_cases',
null_target_procedure=rp.DELETE_ROW,
null_procedure=rp.MEAN)
rp.showInfoDF(weeks, features, n_nulls)
#divide into iq and sj
# get a list of city names
cityNames = weeks['city'].unique().tolist()
iqWeeks = weeks.loc[weeks.city == 'iq']
sjWeeks = weeks.loc[weeks.city == 'sj']
#Outlier cleaning done in Task5
#0 Delete outliers
#1st iteration
iqWeeks.drop(iqWeeks.index[[244, 104, 3, 103, 51, 306, 10, 115, 273]],
inplace=True)