How to use _greater_than method in pandera

Best Python code snippet using pandera_python

currency_indicator_backtest.py

Source:currency_indicator_backtest.py

1import pandas as pd2import numpy as np3import sys4import matplotlib.pyplot as plt5oandapyV20.endpoints.accounts.AccountInstruments6sys.path.insert(1, '/Users/user/Desktop/diff')7from libraries.oanda import get_candles_bid_close8"""9Backtesting currencies and Indicators and Look for major trends:10        11    1.  Call candles for over multiple years12    2.  Calculate Currencies and Indicator.13    3.  go through Data and Graph.14    4.  Report Findings.15    5.  Save in convenient fashion for reuse - answerting different questions.16    6.  Save pulled candles onto disk so do not need to call Oanda every time.17Questions:18    19    What is volume in candle?20"""21# Candle Parameters22_from = '2019-06-01T00:00:00Z'23_to = '2020-01-01T00:00:00Z'24granularity = 'D'25# Pairs26pairs_index = ['AUD_CAD', 'AUD_CHF', 'AUD_HKD', 'AUD_JPY', 'AUD_NZD',27                'AUD_USD', 'CAD_CHF', 'CAD_HKD', 'CAD_JPY', 'CHF_HKD',28                'CHF_JPY', 'EUR_AUD', 'EUR_CAD', 'EUR_CHF', 'EUR_GBP',29                'EUR_HKD', 'EUR_JPY', 'EUR_NZD', 'EUR_USD', 'GBP_AUD',30                'GBP_CAD', 'GBP_CHF', 'GBP_HKD', 'GBP_JPY', 'GBP_NZD',31                'GBP_USD', 'HKD_JPY', 'NZD_CAD', 'NZD_CHF', 'NZD_HKD',32                'NZD_JPY', 'NZD_USD', 'USD_CAD', 'USD_CHF', 'USD_HKD',33                'USD_JPY']34# Currencies35currencies_index = list(set('_'.join(pairs_index).split('_')))36currencies_index.sort()37# Call Candles and assemble in to appropriate Dataframe38##############################################################################39if True:40    pairs = get_candles_bid_close(pairs_index[0], granularity, _from, _to)41    pairs = pairs.drop('volume', axis=1)42    pairs = pairs.set_index('timestamp', drop=True)43    for pair in pairs_index[1:]:44        df = get_candles_bid_close(pair, granularity, _from, _to)45        df = df.set_index('timestamp', drop=True)46        df = df.drop('volume', axis=1)47        pairs = pairs.join(df, how='inner')48    # Fill Forward missing prices, drop top row if any nans remain49    pairs = pairs.fillna(method='ffill').dropna()50# Calculate Currencies and Indicator for each timestamp51##############################################################################52if True:53    def conversion_dict(pairs, currencies):54        '''55        Calculate currencies based on pairs56        Return Currency dictionary based on full set of instrument prices57        Input: pairs:      dictionary:       currenct value of each pair 58               currencies: dictionary:       values can be zero59        '''60        for currency in currencies.keys():61            # If pair contains currency, add it to the subset, making sure that62            # It is added with the currency as the denominator63            pair_subset = []64            for pair in list(pairs.keys()):65                if currency == pair.split('_')[0]:66                    pair_subset.append(1 / pairs[pair])67                if currency == pair.split('_')[1]:68                    pair_subset.append(    pairs[pair])69            currencies[currency] = 1 / (np.array(pair_subset).sum() + 1 )70        return currencies 71    72    currency_dict = dict(zip(currencies_index, [0] * len(currencies_index)))73    currencies = pd.DataFrame(columns = currencies_index)74    for index, row in pairs.iterrows():75        p = row.to_dict()76        currencies.loc[index] = conversion_dict(p, currency_dict)77        78 79    # Rcalculate Pairs based on currency prices80    calculated = pairs.copy()81    for column in calculated:82        left = column.split('_')[0]83        right = column.split('_')[1]84        calculated[column] = currencies[left] / currencies[right]85        86    # Calculate Difference Indicator   87    differences = calculated - pairs88    89            90# Difference test 191##############################################################################92if False:93    '''94    for a currency:95        Bin amount differences were positive 96        for each bin:97            how many times did it happen?98            what was the percent of time that the next days price was higher?99        100    '''101    102    # Only choose those days where the difference was greater than zero103    104    # This is wrong.105    106    '''107    curr = 'EUR_USD'108    greater_than_0_index = differences[curr].values > 0109    diff_greater = differences[curr].values[greater_than_0_index]110    pair_greater = pairs[curr].values[greater_than_0_index]111    bins = pd.cut(diff_greater, 50, labels=False)112    113    114    _bins = []115    _count = []116    _greater_than = []117    for i in range(bins.max() + 1):118        _bins.append(i)119        _count.append((bins == i).sum())120        # greater_than = (pairs[curr].values[bins == i] < \121        #                 pairs[curr].values[np.roll(bins == i, 1)]).mean()122        greater_than = (pair_greater[bins == i] < \123                        pair_greater[np.roll(bins == i, 1)]).mean()124        _greater_than.append(greater_than)125        126    plt.scatter(_count, _greater_than)127    '''128    129    130    131    132    # WY DOES E_USD RETURN ON 300 RESUTLS  ?? ? ? 133    # ANYWAY - TIRED NOW.134    135    136    137    curr = 'EUR_USD'138    bins = pd.cut(differences[curr], 50, labels=False)139    greater_than_0_index = calculated[curr] > pairs[curr]140    diff_greater = differences[curr].values[greater_than_0_index]141    pair_greater = pairs[curr].values[greater_than_0_index]142    # Simple question - if it was higher, was the next ( what about the last)143    _greater = pairs[curr][greater_than_0_index].values < pairs[curr][np.roll(greater_than_0_index, 1)].values144    print(_greater.mean())145    146    # The answer is no - what about if we bin the sizes147    148    bins = pd.cut(differences[curr][greater_than_0_index].values, 50, labels=False)149    150    151    152    _bins = []153    _count = []154    _greater_than = []155    for i in range(bins.max() + 1):156        _bins.append(i)157        _count.append((bins == i).sum())158        # greater_than = (pairs[curr].values[bins == i] < \159        #                 pairs[curr].values[np.roll(bins == i, 1)]).mean()160        greater_than = (pair_greater[bins == i] < \161                        pair_greater[np.roll(bins == i, 1)]).mean()162        _greater_than.append(greater_than)163        164    plt.scatter(_count, _greater_than)165    ...

whole_currnecy_universe_test.py

Source:whole_currnecy_universe_test.py

1import pandas as pd2import numpy as np3import sys4import matplotlib.pyplot as plt5sys.path.insert(1, '/Users/user/Desktop/diff')6from libraries.oanda import get_candles_bid_close7"""8Backtesting currencies and Indicators and Look for major trends:9        10    1.  Call candles for over multiple years11    2.  Calculate Currencies and Indicator.12    3.  go through Data and Graph.13    4.  Report Findings.14    5.  Save in convenient fashion for reuse - answerting different questions.15    6.  Save pulled candles onto disk so do not need to call Oanda every time.16Questions:17    18    What is volume in candle?19"""20# Candle Parameters21_from = '2019-06-01T00:00:00Z'22_to = '2020-01-01T00:00:00Z'23granularity = 'D'24# Pairs25pairs_index = ['AUD_CAD', 'AUD_CHF', 'AUD_HKD', 'AUD_JPY', 'AUD_NZD',26                'AUD_USD', 'CAD_CHF', 'CAD_HKD', 'CAD_JPY', 'CHF_HKD',27                'CHF_JPY', 'EUR_AUD', 'EUR_CAD', 'EUR_CHF', 'EUR_GBP',28                'EUR_HKD', 'EUR_JPY', 'EUR_NZD', 'EUR_USD', 'GBP_AUD',29                'GBP_CAD', 'GBP_CHF', 'GBP_HKD', 'GBP_JPY', 'GBP_NZD',30                'GBP_USD', 'HKD_JPY', 'NZD_CAD', 'NZD_CHF', 'NZD_HKD',31                'NZD_JPY', 'NZD_USD', 'USD_CAD', 'USD_CHF', 'USD_HKD',32                'USD_JPY']33# Currencies34currencies_index = list(set('_'.join(pairs_index).split('_')))35currencies_index.sort()36# Call Candles and assemble in to appropriate Dataframe37##############################################################################38if True:39    pairs = get_candles_bid_close(pairs_index[0], granularity, _from, _to)40    pairs = pairs.drop('volume', axis=1)41    pairs = pairs.set_index('timestamp', drop=True)42    for pair in pairs_index[1:]:43        df = get_candles_bid_close(pair, granularity, _from, _to)44        df = df.set_index('timestamp', drop=True)45        df = df.drop('volume', axis=1)46        pairs = pairs.join(df, how='inner')47    # Fill Forward missing prices, drop top row if any nans remain48    pairs = pairs.fillna(method='ffill').dropna()49# Calculate Currencies and Indicator for each timestamp50##############################################################################51if True:52    def conversion_dict(pairs, currencies):53        '''54        Calculate currencies based on pairs55        Return Currency dictionary based on full set of instrument prices56        Input: pairs:      dictionary:       currenct value of each pair 57               currencies: dictionary:       values can be zero58        '''59        for currency in currencies.keys():60            # If pair contains currency, add it to the subset, making sure that61            # It is added with the currency as the denominator62            pair_subset = []63            for pair in list(pairs.keys()):64                if currency == pair.split('_')[0]:65                    pair_subset.append(1 / pairs[pair])66                if currency == pair.split('_')[1]:67                    pair_subset.append(    pairs[pair])68            currencies[currency] = 1 / (np.array(pair_subset).sum() + 1 )69        return currencies 70    71    currency_dict = dict(zip(currencies_index, [0] * len(currencies_index)))72    currencies = pd.DataFrame(columns = currencies_index)73    for index, row in pairs.iterrows():74        p = row.to_dict()75        currencies.loc[index] = conversion_dict(p, currency_dict)76        77 78    # Rcalculate Pairs based on currency prices79    calculated = pairs.copy()80    for column in calculated:81        left = column.split('_')[0]82        right = column.split('_')[1]83        calculated[column] = currencies[left] / currencies[right]84        85    # Calculate Difference Indicator   86    differences = calculated - pairs87    88            89# Difference test 190##############################################################################91if False:92    '''93    for a currency:94        Bin amount differences were positive 95        for each bin:96            how many times did it happen?97            what was the percent of time that the next days price was higher?98        99    '''100    101    # Only choose those days where the difference was greater than zero102    103    # This is wrong.104    105    '''106    curr = 'EUR_USD'107    greater_than_0_index = differences[curr].values > 0108    diff_greater = differences[curr].values[greater_than_0_index]109    pair_greater = pairs[curr].values[greater_than_0_index]110    bins = pd.cut(diff_greater, 50, labels=False)111    112    113    _bins = []114    _count = []115    _greater_than = []116    for i in range(bins.max() + 1):117        _bins.append(i)118        _count.append((bins == i).sum())119        # greater_than = (pairs[curr].values[bins == i] < \120        #                 pairs[curr].values[np.roll(bins == i, 1)]).mean()121        greater_than = (pair_greater[bins == i] < \122                        pair_greater[np.roll(bins == i, 1)]).mean()123        _greater_than.append(greater_than)124        125    plt.scatter(_count, _greater_than)126    '''127    128    129    130    131    # WY DOES E_USD RETURN ON 300 RESUTLS  ?? ? ? 132    # ANYWAY - TIRED NOW.133    134    135    136    curr = 'EUR_USD'137    bins = pd.cut(differences[curr], 50, labels=False)138    greater_than_0_index = calculated[curr] > pairs[curr]139    diff_greater = differences[curr].values[greater_than_0_index]140    pair_greater = pairs[curr].values[greater_than_0_index]141    # Simple question - if it was higher, was the next ( what about the last)142    _greater = pairs[curr][greater_than_0_index].values < pairs[curr][np.roll(greater_than_0_index, 1)].values143    print(_greater.mean())144    145    # The answer is no - what about if we bin the sizes146    147    bins = pd.cut(differences[curr][greater_than_0_index].values, 50, labels=False)148    149    150    151    _bins = []152    _count = []153    _greater_than = []154    for i in range(bins.max() + 1):155        _bins.append(i)156        _count.append((bins == i).sum())157        # greater_than = (pairs[curr].values[bins == i] < \158        #                 pairs[curr].values[np.roll(bins == i, 1)]).mean()159        greater_than = (pair_greater[bins == i] < \160                        pair_greater[np.roll(bins == i, 1)]).mean()161        _greater_than.append(greater_than)162        163    plt.scatter(_count, _greater_than)164    ...

ranker.py

Source:ranker.py

1#! /usr/bin/env python2import sys3from functools import cmp_to_key4from itertools import combinations5from pyrsistent import pset, pmap_field, PSet, pset_field, PClass, field6class Ranker(PClass):7    _greater_than = pmap_field(str, PSet)8    _remaining_pairs = pset_field(PSet)9    _all_items = field(initial=pset())10    @classmethod11    def new(cls, items):12        remaining = pset({pset(x) for x in combinations(items, 2)})13        return cls(_remaining_pairs=remaining)14    def is_complete(self):15        return len(self._remaining_pairs) == 016    def add_ranking(self, larger, smaller):17        """ Decide that 'larger' is greater than 'smaller' """18        key = pset([larger, smaller])19        slf = self20        if slf.cmp_items(smaller, larger) > 0:21            raise ValueError(f"{smaller} is already greater than {larger}")22        if key not in slf._remaining_pairs:23            return slf24        # Set larger as being greater than smaller25        slf = slf.set(_remaining_pairs=slf._remaining_pairs.remove(key))26        if larger not in slf._greater_than:27            slf = slf.set(_greater_than=slf._greater_than.set(larger, pset()))28        slf = slf.set(29            _greater_than=slf._greater_than.set(larger, slf._greater_than[larger].add(smaller)),30            _all_items=slf._all_items.add(larger).add(smaller),31        )32        # Set larger as being greater than everything which smaller is greater than33        for item in slf.everying_less_than(larger):34            slf = slf.add_ranking(larger, item)35        return slf36    def sample(self):37        for item in self._remaining_pairs:38            return item39    def everying_less_than(self, item):40        return self._greater_than[item]41    def has_item(self, item):42        return item in self._all_items43    def sorted_items(self, *args, **kwargs):44        return self.sorted(self._all_items, *args, **kwargs)45    def sorted(self, items, *args, **kwargs):46        kwargs["key"] = self.key_func47        return sorted(items, *args, **kwargs)48    @property49    def key_func(self):50        return cmp_to_key(self.cmp_items)51    def cmp_items(self, a, b):52        if b in self._greater_than.get(a, {}):53            return 154        if a in self._greater_than.get(b, {}):55            return -156        return 057class Undo(Exception):58    """ Raised when we want to stop """59def pick_one(a, b):60    print("Which is greater?")61    print(f"1. {a}")62    print(f"2. {b}")63    while True:64        result = input("> ")65        if result == "1":66            return a, b67        if result == "2":68            return b, a69        if result == "u":70            raise Undo()71def main():72    items = set(sys.argv[1:])73    if not items:74        print("Warning: No items specified", file=sys.stderr)75        return76    r = Ranker.new(items)77    rankings = []78    s = r.sample()79    while s:80        print()81        print("List:", ", ".join(r.sorted_items(reverse=True)))82        print()83        try:84            larger, smaller = pick_one(*list(s))85        except KeyboardInterrupt:86            break87        except Undo:88            try:89                r = rankings.pop()90            except IndexError:91                pass92            continue93        rankings.append(r)94        r = r.add_ranking(larger, smaller)95        s = r.sample()96    if not r.is_complete():97        print("Warning: Ranking is incomplete")98    for x in r.sorted_items(reverse=True):99        print(x)100if __name__ == "__main__":...

Automation Testing Tutorials

Learn to execute automation testing from scratch with LambdaTest Learning Hub. Right from setting up the prerequisites to run your first automation test, to following best practices and diving deeper into advanced test scenarios. LambdaTest Learning Hubs compile a list of step-by-step guides to help you be proficient with different test automation frameworks i.e. Selenium, Cypress, TestNG etc.