yasha/informer_model
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity

add evaluation for 1 min data

Browse Source
This commit is contained in:
Filip Stefaniuk 2025-04-13 09:47:33 +02:00
parent 030618f589
commit 0ded87a0fb
4 changed files with 765 additions and 245 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

507
notebooks/btcusdt_1m_evaluation.ipynb Normal file
View File

@ -0,0 +1,507 @@
{
"cells": [
{
"cell_type": "code",
"execution_count": 1,
"metadata": {},
"outputs": [],
"source": [
"import pandas as pd\n",
"import pickle\n",
"import plotly.graph_objs as go\n",
"import latextable\n",
"from texttable import Texttable\n",
"from strategy.strategy import (\n",
" BuyAndHoldStrategy,\n",
" MACDStrategy,\n",
" RSIStrategy,\n",
" ModelQuantilePredictionsStrategy,\n",
" ModelGmadlPredictionsStrategy,\n",
" ConcatenatedStrategies\n",
")\n",
"from strategy.util import (\n",
" get_data_windows,\n",
" get_sweep_window_predictions,\n",
" get_predictions_dataframe\n",
")\n",
"from strategy.evaluation import (\n",
" parameter_sweep,\n",
" evaluate_strategy\n",
")\n",
"from strategy.plotting import (\n",
" plot_sweep_results\n",
")\n",
"\n",
"PADDING=5000\n",
"VALID_PART=0.2\n",
"INTERVAL='min'\n",
"METRIC='mod_ir'\n",
"TOP_N=10"
]
},
{
"cell_type": "code",
"execution_count": 2,
"metadata": {},
"outputs": [
{
"name": "stderr",
"output_type": "stream",
"text": [
"Failed to detect the name of this notebook, you can set it manually with the WANDB_NOTEBOOK_NAME environment variable to enable code saving.\n",
"\u001b[34m\u001b[1mwandb\u001b[0m: Downloading large artifact btc-usdt-1m:latest, 3717.80MB. 12 files... \n",
"\u001b[34m\u001b[1mwandb\u001b[0m: 12 of 12 files downloaded. \n",
"Done. 0:0:4.7\n"
]
}
],
"source": [
"data_windows = get_data_windows(\n",
" 'wne-masters-thesis-testing',\n",
" 'btc-usdt-1m:latest',\n",
" min_window=0, \n",
" max_window=5\n",
")"
]
},
{
"cell_type": "code",
"execution_count": 3,
"metadata": {},
"outputs": [],
"source": [
"def sweeps_on_all_windows(data_windows, strategy_class, params, **kwargs):\n",
" result = []\n",
" for in_sample, _ in data_windows:\n",
" data_part = int((1 - VALID_PART) * len(in_sample))\n",
" result.append(parameter_sweep(in_sample[data_part-PADDING:], strategy_class, params, padding=PADDING, interval=INTERVAL, **kwargs))\n",
" return result"
]
},
{
"cell_type": "code",
"execution_count": 4,
"metadata": {},
"outputs": [],
"source": [
"buyandhold_best_strategies = [BuyAndHoldStrategy() for _ in data_windows] "
]
},
{
"cell_type": "code",
"execution_count": 5,
"metadata": {},
"outputs": [
{
"name": "stderr",
"output_type": "stream",
"text": [
"\u001b[34m\u001b[1mwandb\u001b[0m: 2 of 2 files downloaded. \n",
"\u001b[34m\u001b[1mwandb\u001b[0m: 2 of 2 files downloaded. \n",
"\u001b[34m\u001b[1mwandb\u001b[0m: 2 of 2 files downloaded. \n",
"\u001b[34m\u001b[1mwandb\u001b[0m: 2 of 2 files downloaded. \n",
"\u001b[34m\u001b[1mwandb\u001b[0m: 2 of 2 files downloaded. \n",
"\u001b[34m\u001b[1mwandb\u001b[0m: 2 of 2 files downloaded. \n",
"\u001b[34m\u001b[1mwandb\u001b[0m: 2 of 2 files downloaded. \n",
"\u001b[34m\u001b[1mwandb\u001b[0m: 2 of 2 files downloaded. \n",
"\u001b[34m\u001b[1mwandb\u001b[0m: 2 of 2 files downloaded. \n",
"\u001b[34m\u001b[1mwandb\u001b[0m: 2 of 2 files downloaded. \n",
"\u001b[34m\u001b[1mwandb\u001b[0m: 2 of 2 files downloaded. \n",
"\u001b[34m\u001b[1mwandb\u001b[0m: 2 of 2 files downloaded. \n"
]
}
],
"source": [
"# Model with gmadl loss\n",
"SWEEP_ID = 'filipstefaniuk/wne-masters-thesis-testing/s8goxcbz'\n",
"# SWEEP_ID = 'filipstefaniuk/wne-masters-thesis-testing/v3epl3qk'\n",
"# train_gmadl_pred_windows = get_sweep_window_predictions(SWEEP_ID, 'train')\n",
"valid_gmadl_pred_windows = get_sweep_window_predictions(SWEEP_ID, 'valid')\n",
"test_gmadl_pred_windows = get_sweep_window_predictions(SWEEP_ID, 'test')"
]
},
{
"cell_type": "code",
"execution_count": 8,
"metadata": {},
"outputs": [],
"source": [
"# y = test_gmadl_pred_windows[0][2][:, 0, 0]\n",
"# fig = go.Figure([\n",
"# go.Scatter(y=y[::100]),\n",
"# ])\n",
"# fig.show()"
]
},
{
"cell_type": "code",
"execution_count": 11,
"metadata": {},
"outputs": [
{
"name": "stderr",
"output_type": "stream",
"text": [
"100%|██████████| 1176/1176 [04:40<00:00, 4.20it/s]\n",
"100%|██████████| 1176/1176 [04:40<00:00, 4.20it/s]\n",
"100%|██████████| 1176/1176 [04:36<00:00, 4.26it/s]\n",
"100%|██████████| 1176/1176 [04:35<00:00, 4.28it/s]\n",
"100%|██████████| 1176/1176 [04:36<00:00, 4.26it/s]\n",
"100%|██████████| 1176/1176 [04:30<00:00, 4.35it/s]\n"
]
}
],
"source": [
"MODEL_GMADL_LOSS_FILTER = lambda p: (\n",
" ((p['enter_long'] is not None and (p['enter_short'] is not None or p['exit_long'] is not None))\n",
" or (p['enter_short'] is not None and (p['exit_short'] is not None or p['enter_long'] is not None)))\n",
" and (p['enter_short'] is None or p['exit_long'] is None or (p['exit_long'] > p['enter_short']))\n",
" and (p['enter_long'] is None or p['exit_short'] is None or (p['exit_short'] < p['enter_long'])))\n",
"\n",
"gmadl_model_sweep_results = []\n",
"for (in_sample, _), valid_preds, test_preds in zip(data_windows, valid_gmadl_pred_windows, test_gmadl_pred_windows):\n",
" data_part = int((1 - VALID_PART) * len(in_sample))\n",
" params={\n",
" 'predictions': [get_predictions_dataframe(valid_preds, test_preds)],\n",
" 'enter_long': [None, 0.001, 0.002, 0.003, 0.004, 0.005, 0.006, 0.007],\n",
" 'exit_long': [None, -0.001, -0.002, -0.003, -0.004, -0.005, -0.006, -0.007],\n",
" 'enter_short': [None, -0.001, -0.002, -0.003, -0.004, -0.005, -0.006, -0.007],\n",
" 'exit_short': [None, 0.001, 0.002, 0.003, 0.004, 0.005, 0.006, 0.007],\n",
" }\n",
" \n",
" gmadl_model_sweep_results.append(parameter_sweep(\n",
" in_sample[data_part-PADDING:],\n",
" ModelGmadlPredictionsStrategy,\n",
" params,\n",
" params_filter=MODEL_GMADL_LOSS_FILTER,\n",
" padding=PADDING,\n",
" interval=INTERVAL,\n",
" sort_by=METRIC))\n",
" \n",
"\n",
"gmadl_model_best_strategies = [[strat for _, strat in results[:TOP_N]] for results in gmadl_model_sweep_results]"
]
},
{
"cell_type": "code",
"execution_count": 3,
"metadata": {},
"outputs": [
{
"ename": "NameError",
"evalue": "name 'buyandhold_best_strategies' is not defined",
"output_type": "error",
"traceback": [
"\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
"\u001b[0;31mNameError\u001b[0m Traceback (most recent call last)",
"Cell \u001b[0;32mIn[3], line 3\u001b[0m\n\u001b[1;32m 1\u001b[0m \u001b[38;5;66;03m# Persist best strategies, so that they don't have to be recomputed every time\u001b[39;00m\n\u001b[1;32m 2\u001b[0m best_strategies \u001b[38;5;241m=\u001b[39m {\n\u001b[0;32m----> 3\u001b[0m \u001b[38;5;124m'\u001b[39m\u001b[38;5;124mbuy_and_hold\u001b[39m\u001b[38;5;124m'\u001b[39m: \u001b[43mbuyandhold_best_strategies\u001b[49m,\n\u001b[1;32m 4\u001b[0m \u001b[38;5;124m'\u001b[39m\u001b[38;5;124mgmadl_model\u001b[39m\u001b[38;5;124m'\u001b[39m: gmadl_model_best_strategies\n\u001b[1;32m 5\u001b[0m }\n\u001b[1;32m 7\u001b[0m \u001b[38;5;28;01mwith\u001b[39;00m \u001b[38;5;28mopen\u001b[39m(\u001b[38;5;124m'\u001b[39m\u001b[38;5;124mcache/1min-best-strategies-v1.pkl\u001b[39m\u001b[38;5;124m'\u001b[39m, \u001b[38;5;124m'\u001b[39m\u001b[38;5;124mwb\u001b[39m\u001b[38;5;124m'\u001b[39m) \u001b[38;5;28;01mas\u001b[39;00m outp:\n\u001b[1;32m 8\u001b[0m pickle\u001b[38;5;241m.\u001b[39mdump(best_strategies, outp, pickle\u001b[38;5;241m.\u001b[39mHIGHEST_PROTOCOL)\n",
"\u001b[0;31mNameError\u001b[0m: name 'buyandhold_best_strategies' is not defined"
]
}
],
"source": [
"# Persist best strategies, so that they don't have to be recomputed every time\n",
"best_strategies = {\n",
" 'buy_and_hold': buyandhold_best_strategies,\n",
" 'gmadl_model': gmadl_model_best_strategies\n",
"}\n",
"\n",
"with open('cache/1min-best-strategies-v1.pkl', 'wb') as outp:\n",
" pickle.dump(best_strategies, outp, pickle.HIGHEST_PROTOCOL)"
]
},
{
"cell_type": "code",
"execution_count": 4,
"metadata": {},
"outputs": [],
"source": [
"with open('cache/1min-best-strategies-v1.pkl', 'rb') as inpt:\n",
" best_strategies = pickle.load(inpt)"
]
},
{
"cell_type": "code",
"execution_count": 3,
"metadata": {},
"outputs": [],
"source": [
"# plot_sweep_results(pd.DataFrame([result for result, _ in gmadl_model_sweep_results[0]]), parameters=['enter_long', 'exit_long', 'enter_short', 'exit_short'], round=5, objective='mod_ir')"
]
},
{
"cell_type": "code",
"execution_count": 5,
"metadata": {},
"outputs": [],
"source": [
"def results_plot(idx, result_buyandhold, result_gmadl_model, width=850, height=500, notitle=False, v_lines=None):\n",
"\n",
" fig = go.Figure([\n",
" go.Scatter(y=result_buyandhold['portfolio_value'], x=result_buyandhold['time'], name=\"Buy and Hold\"),\n",
" go.Scatter(y=result_gmadl_model['portfolio_value'], x=result_gmadl_model['time'], name='GMADL Informer Strategy')\n",
" ])\n",
" \n",
" if v_lines:\n",
" for v_line in v_lines:\n",
" fig.add_shape(\n",
" go.layout.Shape(type=\"line\",\n",
" yref=\"paper\",\n",
" xref=\"x\",\n",
" x0=v_line,\n",
" x1=v_line,\n",
" y0=0,\n",
" y1=1,\n",
" line=dict(dash='dash', color='rgb(140,140,140)')))\n",
" fig.update_layout(\n",
" title={\n",
" 'text': f\"W{idx}-{INTERVAL}\",\n",
" 'y':0.97,\n",
" 'x':0.5,\n",
" 'xanchor': 'center',\n",
" 'yanchor': 'top'} if not notitle else None,\n",
" yaxis_title=\"Portfolio Value\",\n",
" xaxis_title=\"Date\",\n",
" font=dict(\n",
" # family=\"Courier New, monospace\",\n",
" size=14,\n",
" ),\n",
" autosize=False,\n",
" width=width,\n",
" height=height,\n",
" margin=dict(l=20, r=20, t=20 if notitle else 110, b=20),\n",
" plot_bgcolor='white',\n",
" legend=dict(\n",
" orientation=\"h\",\n",
" yanchor=\"bottom\",\n",
" y=1.02,\n",
" xanchor=\"left\",\n",
" x=0.02\n",
" )\n",
" )\n",
" fig.update_xaxes(\n",
" mirror=True,\n",
" ticks='outside',\n",
" showline=True,\n",
" linecolor='black',\n",
" gridcolor='lightgrey'\n",
" )\n",
" fig.update_yaxes(\n",
" mirror=True,\n",
" ticks='outside',\n",
" showline=True,\n",
" linecolor='black',\n",
" gridcolor='lightgrey'\n",
" )\n",
" fig.write_image(f\"images/eval-w{idx}-{INTERVAL}.png\")\n",
" fig.show()\n",
" \n",
"def results_table(result_buyandhold, result_gmadl_model):\n",
" table_eval_windows = Texttable()\n",
" table_eval_windows.set_deco(Texttable.HEADER)\n",
" table_eval_windows.set_cols_align([\"l\", \"c\",\"c\", \"c\", \"c\", \"c\", \"c\", \"c\", \"c\", \"c\"])\n",
" table_eval_windows.set_precision(3)\n",
"\n",
" table_eval_windows.header([\n",
" \"\\\\textbf{Strategy}\",\n",
" \"\\\\textbf{VAL}\",\n",
" \"\\\\textbf{ARC}\",\n",
" \"\\\\textbf{ASD}\",\n",
" \"\\\\textbf{IR*}\",\n",
" \"\\\\textbf{MD}\",\n",
" \"\\\\textbf{IR**}\",\n",
" \"\\\\textbf{N}\",\n",
" \"\\\\textbf{LONG}\",\n",
" \"\\\\textbf{SHORT}\",\n",
" ])\n",
"\n",
" strategy_name_result = [\n",
" ('Buy and Hold', result_buyandhold),\n",
" ('GMADL Informer', result_gmadl_model)\n",
" ]\n",
" for strategy_name, result in strategy_name_result:\n",
" table_eval_windows.add_row([\n",
" strategy_name,\n",
" result['value'],\n",
" f\"{result['arc']*100:.2f}\\%\",\n",
" f\"{result['asd']*100:.2f}\\%\",\n",
" result['ir'],\n",
" f\"{result['md']*100:.2f}\\%\",\n",
" result['mod_ir'],\n",
" result['n_trades'],\n",
" f\"{result['long_pos']*100:.2f}\\%\",\n",
" f\"{result['short_pos']*100:.2f}\\%\",\n",
" ])\n",
" print(latextable.draw_latex(table_eval_windows))\n"
]
},
{
"cell_type": "code",
"execution_count": 6,
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"\\begin{table}\n",
"\t\\begin{center}\n",
"\t\t\\begin{tabular}{lccccccccc}\n",
"\t\t\t\\textbf{Strategy} & \\textbf{VAL} & \\textbf{ARC} & \\textbf{ASD} & \\textbf{IR*} & \\textbf{MD} & \\textbf{IR**} & \\textbf{N} & \\textbf{LONG} & \\textbf{SHORT} \\\\\n",
"\t\t\t\\hline\n",
"\t\t\tBuy and Hold & 0.929 & -13.87\\% & 69.66\\% & -0.199 & 52.09\\% & -0.053 & 2 & 100.00\\% & 0.00\\% \\\\\n",
"\t\t\tGMADL Informer & 1.306 & 71.83\\% & 69.69\\% & 1.031 & 41.57\\% & 1.781 & 50 & 7.29\\% & 92.71\\% \\\\\n",
"\t\t\\end{tabular}\n",
"\t\\end{center}\n",
"\\end{table}\n",
"\\begin{table}\n",
"\t\\begin{center}\n",
"\t\t\\begin{tabular}{lccccccccc}\n",
"\t\t\t\\textbf{Strategy} & \\textbf{VAL} & \\textbf{ARC} & \\textbf{ASD} & \\textbf{IR*} & \\textbf{MD} & \\textbf{IR**} & \\textbf{N} & \\textbf{LONG} & \\textbf{SHORT} \\\\\n",
"\t\t\t\\hline\n",
"\t\t\tBuy and Hold & 0.549 & -70.35\\% & 73.36\\% & -0.959 & 63.40\\% & -1.064 & 2 & 100.00\\% & 0.00\\% \\\\\n",
"\t\t\tGMADL Informer & 1.837 & 243.15\\% & 73.38\\% & 3.314 & 25.16\\% & 32.024 & 186 & 18.19\\% & 81.81\\% \\\\\n",
"\t\t\\end{tabular}\n",
"\t\\end{center}\n",
"\\end{table}\n",
"\\begin{table}\n",
"\t\\begin{center}\n",
"\t\t\\begin{tabular}{lccccccccc}\n",
"\t\t\t\\textbf{Strategy} & \\textbf{VAL} & \\textbf{ARC} & \\textbf{ASD} & \\textbf{IR*} & \\textbf{MD} & \\textbf{IR**} & \\textbf{N} & \\textbf{LONG} & \\textbf{SHORT} \\\\\n",
"\t\t\t\\hline\n",
"\t\t\tBuy and Hold & 1.016 & 3.33\\% & 52.45\\% & 0.064 & 38.42\\% & 0.006 & 2 & 100.00\\% & 0.00\\% \\\\\n",
"\t\t\tGMADL Informer & 0.739 & -45.82\\% & 52.21\\% & -0.878 & 42.46\\% & -0.947 & 35 & 4.70\\% & 93.05\\% \\\\\n",
"\t\t\\end{tabular}\n",
"\t\\end{center}\n",
"\\end{table}\n",
"\\begin{table}\n",
"\t\\begin{center}\n",
"\t\t\\begin{tabular}{lccccccccc}\n",
"\t\t\t\\textbf{Strategy} & \\textbf{VAL} & \\textbf{ARC} & \\textbf{ASD} & \\textbf{IR*} & \\textbf{MD} & \\textbf{IR**} & \\textbf{N} & \\textbf{LONG} & \\textbf{SHORT} \\\\\n",
"\t\t\t\\hline\n",
"\t\t\tBuy and Hold & 1.230 & 52.29\\% & 44.30\\% & 1.180 & 22.35\\% & 2.761 & 2 & 100.00\\% & 0.00\\% \\\\\n",
"\t\t\tGMADL Informer & 1.086 & 18.12\\% & 40.58\\% & 0.446 & 26.30\\% & 0.308 & 11 & 60.03\\% & 23.82\\% \\\\\n",
"\t\t\\end{tabular}\n",
"\t\\end{center}\n",
"\\end{table}\n",
"\\begin{table}\n",
"\t\\begin{center}\n",
"\t\t\\begin{tabular}{lccccccccc}\n",
"\t\t\t\\textbf{Strategy} & \\textbf{VAL} & \\textbf{ARC} & \\textbf{ASD} & \\textbf{IR*} & \\textbf{MD} & \\textbf{IR**} & \\textbf{N} & \\textbf{LONG} & \\textbf{SHORT} \\\\\n",
"\t\t\t\\hline\n",
"\t\t\tBuy and Hold & 1.439 & 109.31\\% & 43.75\\% & 2.498 & 21.12\\% & 12.930 & 2 & 100.00\\% & 0.00\\% \\\\\n",
"\t\t\tGMADL Informer & 1.010 & 1.98\\% & 43.47\\% & 0.046 & 31.96\\% & 0.003 & 67 & 80.24\\% & 15.37\\% \\\\\n",
"\t\t\\end{tabular}\n",
"\t\\end{center}\n",
"\\end{table}\n",
"\\begin{table}\n",
"\t\\begin{center}\n",
"\t\t\\begin{tabular}{lccccccccc}\n",
"\t\t\t\\textbf{Strategy} & \\textbf{VAL} & \\textbf{ARC} & \\textbf{ASD} & \\textbf{IR*} & \\textbf{MD} & \\textbf{IR**} & \\textbf{N} & \\textbf{LONG} & \\textbf{SHORT} \\\\\n",
"\t\t\t\\hline\n",
"\t\t\tBuy and Hold & 1.561 & 146.58\\% & 53.72\\% & 2.729 & 27.11\\% & 14.756 & 2 & 100.00\\% & 0.00\\% \\\\\n",
"\t\t\tGMADL Informer & 1.178 & 39.32\\% & 43.06\\% & 0.913 & 18.63\\% & 1.927 & 92 & 54.86\\% & 0.00\\% \\\\\n",
"\t\t\\end{tabular}\n",
"\t\\end{center}\n",
"\\end{table}\n"
]
}
],
"source": [
"for i, (in_sample, out_of_sample) in enumerate(data_windows):\n",
" padded_window = pd.concat([in_sample.iloc[-PADDING:], out_of_sample])\n",
" result_buyandhold = evaluate_strategy(padded_window, best_strategies['buy_and_hold'][i], padding=PADDING, interval=INTERVAL)\n",
" result_gmadl_model = evaluate_strategy(padded_window, [s[0] for s in best_strategies['gmadl_model']][i], padding=PADDING, interval=INTERVAL)\n",
"\n",
" results_table(result_buyandhold, result_gmadl_model)\n",
" # results_plot(i+1, result_buyandhold, result_gmadl_model)"
]
},
{
"cell_type": "code",
"execution_count": 2,
"metadata": {},
"outputs": [],
"source": [
"# test_data = pd.concat([data_windows[0][0][-PADDING:]] + [data_window[1] for data_window in data_windows])\n",
"# buy_and_hold_concat = evaluate_strategy(test_data, BuyAndHoldStrategy(), padding=PADDING, interval=INTERVAL)\n",
"# gmadl_model_concat = evaluate_strategy(test_data, ConcatenatedStrategies(len(data_windows[0][1]), [s[0] for s in best_strategies['gmadl_model']], padding=PADDING), padding=PADDING, interval=INTERVAL)\n",
"\n",
"# v_lines=[data_window[1]['close_time'].iloc[-1] for data_window in data_windows][:-1]\n",
"# results_table(buy_and_hold_concat, gmadl_model_concat)\n",
"# results_plot(0, buy_and_hold_concat, gmadl_model_concat, width=1300, height=500, notitle=True)"
]
},
{
"cell_type": "code",
"execution_count": 3,
"metadata": {},
"outputs": [],
"source": [
"import plotly.figure_factory as ff\n",
"\n",
"def results_for_strats(\n",
" data_windows, \n",
" best_strategies,\n",
" top_n=10):\n",
" test_data = pd.concat([data_windows[0][0][-PADDING:]] + [data_window[1] for data_window in data_windows])\n",
"\n",
" buy_and_hold_concat = evaluate_strategy(test_data, BuyAndHoldStrategy(), padding=PADDING, interval='min')\n",
" gmadl_1min_model_concat = [evaluate_strategy(test_data, ConcatenatedStrategies(len(data_windows[0][1]), [s[x] for s in best_strategies['gmadl_model']], padding=PADDING), padding=PADDING, interval='min') for x in range(top_n)]\n",
"\n",
" z = list(reversed([\n",
" list(reversed([round(buy_and_hold_concat['mod_ir'], 3)]*top_n)),\n",
" list(reversed([round(x['mod_ir'], 3) for x in gmadl_1min_model_concat])),\n",
" ]))\n",
" x = list(reversed(range(1, top_n+1)))\n",
" y = list(reversed([\n",
" \"Buy and Hold\",\n",
" \"Gmadl Informer (1 min)\"\n",
" ]))\n",
" # 'Portland'\n",
" fig = ff.create_annotated_heatmap(z, x=x, y=y, colorscale='thermal', zmid=buy_and_hold_concat['mod_ir'])\n",
" fig.update_layout(\n",
" margin=dict(l=20, r=20, b=20, t=20),\n",
" width=1100,\n",
" height=650,\n",
" font=dict(\n",
" # family=\"Courier New, monospace\",\n",
" size=16, # Set the font size here\n",
" # color=\"RebeccaPurple\"\n",
" )\n",
" )\n",
" fig.show()\n",
"\n",
"# results_for_strats(data_windows, best_strategies, top_n=10) "
]
},
{
"cell_type": "code",
"execution_count": null,
"metadata": {},
"outputs": [],
"source": []
}
],
"metadata": {
"kernelspec": {
"display_name": "wnemsc",
"language": "python",
"name": "python3"
},
"language_info": {
"codemirror_mode": {
"name": "ipython",
"version": 3
},
"file_extension": ".py",
"mimetype": "text/x-python",
"name": "python",
"nbconvert_exporter": "python",
"pygments_lexer": "ipython3",
"version": "3.9.19"
}
},
"nbformat": 4,
"nbformat_minor": 2
}

53
notebooks/btcusdt_5m_evaluation.ipynb
View File

@ -2,7 +2,7 @@
"cells": [ "cells": [
{ {
"cell_type": "code", "cell_type": "code",
"execution_count": 1, "execution_count": 2,
"metadata": {}, "metadata": {},
"outputs": [], "outputs": [],
"source": [ "source": [
@ -41,7 +41,7 @@
}, },
{ {
"cell_type": "code", "cell_type": "code",
"execution_count": 2, "execution_count": 3,
"metadata": {}, "metadata": {},
"outputs": [ "outputs": [
{ {
@ -51,7 +51,7 @@
"Failed to detect the name of this notebook, you can set it manually with the WANDB_NOTEBOOK_NAME environment variable to enable code saving.\n", "Failed to detect the name of this notebook, you can set it manually with the WANDB_NOTEBOOK_NAME environment variable to enable code saving.\n",
"\u001b[34m\u001b[1mwandb\u001b[0m: Downloading large artifact btc-usdt-5m:latest, 745.12MB. 12 files... \n", "\u001b[34m\u001b[1mwandb\u001b[0m: Downloading large artifact btc-usdt-5m:latest, 745.12MB. 12 files... \n",
"\u001b[34m\u001b[1mwandb\u001b[0m: 12 of 12 files downloaded. \n", "\u001b[34m\u001b[1mwandb\u001b[0m: 12 of 12 files downloaded. \n",
"Done. 0:0:1.4\n" "Done. 0:0:1.3\n"
] ]
} }
], ],
@ -369,13 +369,14 @@
}, },
{ {
"cell_type": "code", "cell_type": "code",
"execution_count": 12, "execution_count": 2,
"metadata": {}, "metadata": {},
"outputs": [ "outputs": [
{ {
"name": "stderr", "name": "stderr",
"output_type": "stream", "output_type": "stream",
"text": [ "text": [
"Failed to detect the name of this notebook, you can set it manually with the WANDB_NOTEBOOK_NAME environment variable to enable code saving.\n",
"\u001b[34m\u001b[1mwandb\u001b[0m: 2 of 2 files downloaded. \n", "\u001b[34m\u001b[1mwandb\u001b[0m: 2 of 2 files downloaded. \n",
"\u001b[34m\u001b[1mwandb\u001b[0m: 2 of 2 files downloaded. \n", "\u001b[34m\u001b[1mwandb\u001b[0m: 2 of 2 files downloaded. \n",
"\u001b[34m\u001b[1mwandb\u001b[0m: 2 of 2 files downloaded. \n", "\u001b[34m\u001b[1mwandb\u001b[0m: 2 of 2 files downloaded. \n",
@ -478,7 +479,7 @@
}, },
{ {
"cell_type": "code", "cell_type": "code",
"execution_count": 3, "execution_count": 4,
"metadata": {}, "metadata": {},
"outputs": [], "outputs": [],
"source": [ "source": [
@ -1022,6 +1023,48 @@
"\n" "\n"
] ]
}, },
{
"cell_type": "code",
"execution_count": 7,
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"\\begin{table}\n",
"\t\\begin{center}\n",
"\t\t\\begin{tabular}{lccccccccc}\n",
"\t\t\t\\textbf{Strategy} & \\textbf{VAL} & \\textbf{ARC} & \\textbf{ASD} & \\textbf{IR*} & \\textbf{MD} & \\textbf{IR**} & \\textbf{N} & \\textbf{LONG} & \\textbf{SHORT} \\\\\n",
"\t\t\t\\hline\n",
"\t\t\tBuy and Hold & 1.441 & 13.14\\% & 57.74\\% & 0.228 & 77.31\\% & 0.039 & 2 & 100.00\\% & 0.00\\% \\\\\n",
"\t\t\tMACD Strategy & 1.918 & 24.63\\% & 54.07\\% & 0.456 & 67.19\\% & 0.167 & 2535 & 50.39\\% & 32.38\\% \\\\\n",
"\t\t\tRSI Strategy & 5.154 & 74.05\\% & 50.37\\% & 1.470 & 26.46\\% & 4.113 & 846 & 28.29\\% & 33.47\\% \\\\\n",
"\t\t\tRMSE Informer & 0.650 & -13.53\\% & 15.13\\% & -0.894 & 44.05\\% & -0.275 & 16 & 0.00\\% & 9.58\\% \\\\\n",
"\t\t\tQuantile Informer & 5.404 & 76.86\\% & 47.83\\% & 1.607 & 37.34\\% & 3.307 & 3395 & 40.24\\% & 27.84\\% \\\\\n",
"\t\t\tGMADL Informer & 14.946 & 149.43\\% & 54.42\\% & 2.746 & 31.57\\% & 12.994 & 846 & 44.80\\% & 41.51\\% \\\\\n",
"\t\t\\end{tabular}\n",
"\t\\end{center}\n",
"\\end{table}\n"
]
}
],
"source": [
"# No transaction fees\n",
"test_data = pd.concat([data_windows[0][0][-PADDING:]] + [data_window[1] for data_window in data_windows])\n",
"buy_and_hold_concat = evaluate_strategy(test_data, BuyAndHoldStrategy(), padding=PADDING)\n",
"macd_concat = evaluate_strategy(test_data, ConcatenatedStrategies(len(data_windows[0][1]), [s[0] for s in best_strategies['macd_strategies']], padding=PADDING), padding=PADDING, interval=INTERVAL, exchange_fee=0)\n",
"rsi_concat = evaluate_strategy(test_data, ConcatenatedStrategies(len(data_windows[0][1]), [s[0] for s in best_strategies['rsi_strategies']], padding=PADDING), padding=PADDING, interval=INTERVAL, exchange_fee=0)\n",
"rmse_model_concat = evaluate_strategy(test_data, ConcatenatedStrategies(len(data_windows[0][1]), [s[0] for s in best_strategies['rmse_model']], padding=PADDING), padding=PADDING, interval=INTERVAL, exchange_fee=0)\n",
"quantile_model_concat = evaluate_strategy(test_data, ConcatenatedStrategies(len(data_windows[0][1]), [s[0] for s in best_strategies['quantile_model']], padding=PADDING), padding=PADDING, interval=INTERVAL, exchange_fee=0)\n",
"gmadl_model_concat = evaluate_strategy(test_data, ConcatenatedStrategies(len(data_windows[0][1]), [s[0] for s in best_strategies['gmadl_model']], padding=PADDING), padding=PADDING, interval=INTERVAL, exchange_fee=0)\n",
"\n",
"v_lines=[data_window[1]['close_time'].iloc[-1] for data_window in data_windows][:-1]\n",
"results_table(buy_and_hold_concat, macd_concat, rsi_concat, rmse_model_concat, quantile_model_concat, gmadl_model_concat)\n",
"# results_plot(0, buy_and_hold_concat, macd_concat, rsi_concat, rmse_model_concat, quantile_model_concat, gmadl_model_concat, width=1300, height=500, notitle=True)\n",
"\n"
]
},
{ {
"cell_type": "code", "cell_type": "code",
"execution_count": null, "execution_count": null,

301
notebooks/btcusdt_dataset.ipynb
View File

@ -110,11 +110,22 @@
"metadata": {}, "metadata": {},
"outputs": [], "outputs": [],
"source": [ "source": [
"INTERVAL_1_MIN = pd.Timedelta(minutes=1)\n",
"INTERVAL_5_MIN = pd.Timedelta(minutes=5)\n", "INTERVAL_5_MIN = pd.Timedelta(minutes=5)\n",
"INTERVAL_15_MIN = pd.Timedelta(minutes=15)\n", "INTERVAL_15_MIN = pd.Timedelta(minutes=15)\n",
"INTERVAL_30_MIN = pd.Timedelta(minutes=30)" "INTERVAL_30_MIN = pd.Timedelta(minutes=30)"
] ]
}, },
{
"cell_type": "code",
"execution_count": 1,
"metadata": {},
"outputs": [],
"source": [
"# btc_usdt_1_min_data = load_btc_usdt_dataset('../data/raw_data/btc-usdt-1m.csv', INTERVAL_1_MIN)\n",
"# btc_usdt_1_min_data.head()"
]
},
{ {
"cell_type": "code", "cell_type": "code",
"execution_count": 3, "execution_count": 3,
@ -154,7 +165,7 @@
}, },
{ {
"cell_type": "code", "cell_type": "code",
"execution_count": 66, "execution_count": 7,
"metadata": {}, "metadata": {},
"outputs": [], "outputs": [],
"source": [ "source": [
@ -180,6 +191,18 @@
"vix_data['date'] = pd.to_datetime(vix_data['date'])" "vix_data['date'] = pd.to_datetime(vix_data['date'])"
] ]
}, },
{
"cell_type": "code",
"execution_count": 2,
"metadata": {},
"outputs": [],
"source": [
"# vix_1_min_data = preprocess_vix_data(btc_usdt_1_min_data, vix_data)\n",
"# btc_usdt_1_min_data[VIX_COL_NAME] = vix_1_min_data['close']\n",
"# go.Figure(go.Scatter(y=btc_usdt_1_min_data['vix_close_price'].iloc[::10], x=btc_usdt_1_min_data['close_time'].iloc[::10])).show()\n",
"# btc_usdt_1_min_data.head()"
]
},
{ {
"cell_type": "code", "cell_type": "code",
"execution_count": 4, "execution_count": 4,
@ -225,7 +248,7 @@
}, },
{ {
"cell_type": "code", "cell_type": "code",
"execution_count": 70, "execution_count": 9,
"metadata": {}, "metadata": {},
"outputs": [], "outputs": [],
"source": [ "source": [
@ -251,6 +274,18 @@
"fed_data['date'] = pd.to_datetime(fed_data['date'])" "fed_data['date'] = pd.to_datetime(fed_data['date'])"
] ]
}, },
{
"cell_type": "code",
"execution_count": 3,
"metadata": {},
"outputs": [],
"source": [
"# fed_1_min_data = preprocess_fed_data(btc_usdt_1_min_data, fed_data)\n",
"# btc_usdt_1_min_data[FED_COL_NAME] = fed_1_min_data['fedfunds']\n",
"# go.Figure(go.Scatter(y=btc_usdt_1_min_data['effective_rates'].iloc[::10], x=btc_usdt_1_min_data['close_time'].iloc[::10])).show()\n",
"# btc_usdt_1_min_data.head()"
]
},
{ {
"cell_type": "code", "cell_type": "code",
"execution_count": 7, "execution_count": 7,
@ -296,7 +331,7 @@
}, },
{ {
"cell_type": "code", "cell_type": "code",
"execution_count": 74, "execution_count": 11,
"metadata": {}, "metadata": {},
"outputs": [], "outputs": [],
"source": [ "source": [
@ -323,6 +358,18 @@
"feargreed_data = feargreed_data.sort_values('timestamp').reset_index(drop=True)" "feargreed_data = feargreed_data.sort_values('timestamp').reset_index(drop=True)"
] ]
}, },
{
"cell_type": "code",
"execution_count": 4,
"metadata": {},
"outputs": [],
"source": [
"# feargreed_1_min_data = preprocess_feargreed_data(btc_usdt_1_min_data, feargreed_data)\n",
"# btc_usdt_1_min_data[FEAR_GREED_COL_NAME] = feargreed_1_min_data['value']\n",
"# go.Figure(go.Scatter(y=btc_usdt_1_min_data['fear_greed_index'].iloc[::10], x=btc_usdt_1_min_data['close_time'].iloc[::10])).show()\n",
"# btc_usdt_1_min_data.head()"
]
},
{ {
"cell_type": "code", "cell_type": "code",
"execution_count": 10, "execution_count": 10,
@ -369,7 +416,7 @@
}, },
{ {
"cell_type": "code", "cell_type": "code",
"execution_count": 91, "execution_count": 13,
"metadata": {}, "metadata": {},
"outputs": [], "outputs": [],
"source": [ "source": [
@ -448,13 +495,14 @@
}, },
{ {
"cell_type": "code", "cell_type": "code",
"execution_count": 92, "execution_count": 14,
"metadata": {}, "metadata": {},
"outputs": [], "outputs": [],
"source": [ "source": [
"btc_usdt_5_min_data = preprocess_augment_data(btc_usdt_5_min_data, 5)\n", "btc_usdt_1_min_data = preprocess_augment_data(btc_usdt_1_min_data, 1)\n",
"btc_usdt_15_min_data = preprocess_augment_data(btc_usdt_15_min_data, 15)\n", "# btc_usdt_5_min_data = preprocess_augment_data(btc_usdt_5_min_data, 5)\n",
"btc_usdt_30_min_data = preprocess_augment_data(btc_usdt_30_min_data, 30)\n" "# btc_usdt_15_min_data = preprocess_augment_data(btc_usdt_15_min_data, 15)\n",
"# btc_usdt_30_min_data = preprocess_augment_data(btc_usdt_30_min_data, 30)\n"
] ]
}, },
{ {
@ -466,13 +514,14 @@
}, },
{ {
"cell_type": "code", "cell_type": "code",
"execution_count": 93, "execution_count": 15,
"metadata": {}, "metadata": {},
"outputs": [], "outputs": [],
"source": [ "source": [
"btc_usdt_5_min_data.to_csv('../data/preprocessed_data/processed-btc-usdt-5m.csv', index=False)\n", "btc_usdt_1_min_data.to_csv('../data/preprocessed_data/processed-btc-usdt-1m.csv', index=False)\n",
"btc_usdt_15_min_data.to_csv('../data/preprocessed_data/processed-btc-usdt-15m.csv', index=False)\n", "# btc_usdt_5_min_data.to_csv('../data/preprocessed_data/processed-btc-usdt-5m.csv', index=False)\n",
"btc_usdt_30_min_data.to_csv('../data/preprocessed_data/processed-btc-usdt-30m.csv', index=False)" "# btc_usdt_15_min_data.to_csv('../data/preprocessed_data/processed-btc-usdt-15m.csv', index=False)\n",
"# btc_usdt_30_min_data.to_csv('../data/preprocessed_data/processed-btc-usdt-30m.csv', index=False)"
] ]
}, },
{ {
@ -505,6 +554,39 @@
" print(f\"Variables: {dataset.columns}\")\n" " print(f\"Variables: {dataset.columns}\")\n"
] ]
}, },
{
"cell_type": "code",
"execution_count": 17,
"metadata": {},
"outputs": [
{
"name": "stdout",
"output_type": "stream",
"text": [
"---- DATASET BTC-USDT 1 m ----\n",
"Num observations: 2941560\n",
"First observation: 2019-07-21 00:00:59.999000\n",
"Last observation: 2025-02-21 17:59:59.999000\n",
"Time span: 2042 days (5.6 years)\n",
"Variables: Index(['open_time', 'open_price', 'high_price', 'low_price', 'close_price',\n",
" 'volume', 'close_time', 'vix_close_price', 'effective_rates',\n",
" 'fear_greed_index', 'time_index', 'group_id', 'hour', 'weekday',\n",
" 'open_to_close_price', 'high_to_close_price', 'low_to_close_price',\n",
" 'high_to_low_price', 'returns', 'returns_binary', 'log_returns',\n",
" 'vol_1h', 'sma_1h_to_close_price', 'ema_1h_to_close_price', 'vol_1d',\n",
" 'sma_1d_to_close_price', 'ema_1d_to_close_price', 'vol_7d',\n",
" 'sma_7d_to_close_price', 'macd', 'macd_signal', 'rsi',\n",
" 'low_bband_to_close_price', 'up_bband_to_close_price',\n",
" 'mid_bband_to_close_price'],\n",
" dtype='object')\n"
]
}
],
"source": [
"btc_usdt_1_min_data = load_dataset('../data/preprocessed_data/processed-btc-usdt-1m.csv')\n",
"print_dataset_stats(btc_usdt_1_min_data, 1)"
]
},
{ {
"cell_type": "code", "cell_type": "code",
"execution_count": 3, "execution_count": 3,
@ -613,7 +695,7 @@
}, },
{ {
"cell_type": "code", "cell_type": "code",
"execution_count": 5, "execution_count": 4,
"metadata": {}, "metadata": {},
"outputs": [], "outputs": [],
"source": [ "source": [
@ -648,7 +730,7 @@
}, },
{ {
"cell_type": "code", "cell_type": "code",
"execution_count": 6, "execution_count": 5,
"metadata": {}, "metadata": {},
"outputs": [], "outputs": [],
"source": [ "source": [
@ -697,17 +779,34 @@
}, },
{ {
"cell_type": "code", "cell_type": "code",
"execution_count": 8, "execution_count": 5,
"metadata": {}, "metadata": {},
"outputs": [], "outputs": [],
"source": [ "source": [
"windows_5m_data = split_dataset_moving_window(\n", "# windows_1m_data = split_dataset_moving_window(\n",
" trim_number_of_observations(btc_usdt_5_min_data, limit_time=LIMIT_TIME),\n", "# trim_number_of_observations(btc_usdt_1_min_data, limit_time=LIMIT_TIME),\n",
" NUM_MOVING_WINDOWS,\n", "# NUM_MOVING_WINDOWS,\n",
" in_sample_size=(24 * 30 * 24 * (60 // 5)), # 24 months\n", "# in_sample_size=(24 * 30 * 24 * 60), # 24 months\n",
" out_of_sample_size=(6 * 30 * 24 * (60 // 5)) # 6 months\n", "# out_of_sample_size=(6 * 30 * 24 * 60) # 6 months\n",
")\n", "# )\n",
"plot_moving_windows(btc_usdt_5_min_data, windows_5m_data)" "# plot_moving_windows(btc_usdt_1_min_data, windows_1m_data)"
]
},
{
"cell_type": "code",
"execution_count": 6,
"metadata": {},
"outputs": [],
"source": [
"# windows_5m_data = split_dataset_moving_window(\n",
"# trim_number_of_observations(btc_usdt_5_min_data, limit_time=LIMIT_TIME),\n",
"# NUM_MOVING_WINDOWS,\n",
"# in_sample_size=(24 * 30 * 24 * (60 // 5)), # 24 months\n",
"# out_of_sample_size=(6 * 30 * 24 * (60 // 5)) # 6 months\n",
"# )\n",
"# plot_moving_windows(btc_usdt_5_min_data, windows_5m_data)\n",
"# for window in windows_5m_data:\n",
"# print(window[0]['open_time'][0], window[0]['open_time'].iloc[-1])"
] ]
}, },
{ {
@ -751,7 +850,7 @@
}, },
{ {
"cell_type": "code", "cell_type": "code",
"execution_count": 11, "execution_count": 21,
"metadata": {}, "metadata": {},
"outputs": [], "outputs": [],
"source": [ "source": [
@ -762,7 +861,7 @@
}, },
{ {
"cell_type": "code", "cell_type": "code",
"execution_count": 12, "execution_count": 22,
"metadata": {}, "metadata": {},
"outputs": [], "outputs": [],
"source": [ "source": [
@ -785,156 +884,16 @@
}, },
{ {
"cell_type": "code", "cell_type": "code",
"execution_count": 16, "execution_count": 7,
"metadata": {}, "metadata": {},
"outputs": [ "outputs": [],
{
"data": {
"text/html": [
"Finishing last run (ID:8ybedy92) before initializing another..."
],
"text/plain": [
"<IPython.core.display.HTML object>"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
" View run <strong style=\"color:#cdcd00\">legendary-rose-1822</strong> at: <a href='https://wandb.ai/filipstefaniuk/wne-masters-thesis-testing/runs/8ybedy92' target=\"_blank\">https://wandb.ai/filipstefaniuk/wne-masters-thesis-testing/runs/8ybedy92</a><br/> View project at: <a href='https://wandb.ai/filipstefaniuk/wne-masters-thesis-testing' target=\"_blank\">https://wandb.ai/filipstefaniuk/wne-masters-thesis-testing</a><br/>Synced 6 W&B file(s), 0 media file(s), 6 artifact file(s) and 0 other file(s)"
],
"text/plain": [
"<IPython.core.display.HTML object>"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"Find logs at: <code>./wandb/run-20241101_160026-8ybedy92/logs</code>"
],
"text/plain": [
"<IPython.core.display.HTML object>"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"The new W&B backend becomes opt-out in version 0.18.0; try it out with `wandb.require(\"core\")`! See https://wandb.me/wandb-core for more information."
],
"text/plain": [
"<IPython.core.display.HTML object>"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"Successfully finished last run (ID:8ybedy92). Initializing new run:<br/>"
],
"text/plain": [
"<IPython.core.display.HTML object>"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"wandb version 0.18.5 is available! To upgrade, please run:\n",
" $ pip install wandb --upgrade"
],
"text/plain": [
"<IPython.core.display.HTML object>"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"Tracking run with wandb version 0.17.7"
],
"text/plain": [
"<IPython.core.display.HTML object>"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"Run data is saved locally in <code>/Users/filipstefaniuk/Documents/wne-msc-thesis/notebooks/wandb/run-20241101_160127-dtz0yasv</code>"
],
"text/plain": [
"<IPython.core.display.HTML object>"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
"Syncing run <strong><a href='https://wandb.ai/filipstefaniuk/wne-masters-thesis-testing/runs/dtz0yasv' target=\"_blank\">alight-triumph-1823</a></strong> to <a href='https://wandb.ai/filipstefaniuk/wne-masters-thesis-testing' target=\"_blank\">Weights & Biases</a> (<a href='https://wandb.me/run' target=\"_blank\">docs</a>)<br/>"
],
"text/plain": [
"<IPython.core.display.HTML object>"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
" View project at <a href='https://wandb.ai/filipstefaniuk/wne-masters-thesis-testing' target=\"_blank\">https://wandb.ai/filipstefaniuk/wne-masters-thesis-testing</a>"
],
"text/plain": [
"<IPython.core.display.HTML object>"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"data": {
"text/html": [
" View run at <a href='https://wandb.ai/filipstefaniuk/wne-masters-thesis-testing/runs/dtz0yasv' target=\"_blank\">https://wandb.ai/filipstefaniuk/wne-masters-thesis-testing/runs/dtz0yasv</a>"
],
"text/plain": [
"<IPython.core.display.HTML object>"
]
},
"metadata": {},
"output_type": "display_data"
},
{
"name": "stderr",
"output_type": "stream",
"text": [
"\u001b[34m\u001b[1mwandb\u001b[0m: Adding directory to artifact (/var/folders/pz/gkm59rg174z0867wc4h3wd000000gn/T/tmpxkxgri0n)... Done. 0.6s\n"
]
}
],
"source": [ "source": [
"WANDB_PROJECT = 'wne-masters-thesis-testing'\n", "# WANDB_PROJECT = 'wne-masters-thesis-testing'\n",
"\n", "\n",
"upload_dataset_to_wandb(windows_5m_data, 'btc-usdt-5m', project=WANDB_PROJECT)\n", "# upload_dataset_to_wandb(windows_1m_data, 'btc-usdt-1m', project=WANDB_PROJECT)\n",
"upload_dataset_to_wandb(windows_15m_data, 'btc-usdt-15m', project=WANDB_PROJECT)\n", "# upload_dataset_to_wandb(windows_5m_data, 'btc-usdt-5m', project=WANDB_PROJECT)\n",
"upload_dataset_to_wandb(windows_30m_data, 'btc-usdt-30m', project=WANDB_PROJECT)\n" "# upload_dataset_to_wandb(windows_15m_data, 'btc-usdt-15m', project=WANDB_PROJECT)\n",
"# upload_dataset_to_wandb(windows_30m_data, 'btc-usdt-30m', project=WANDB_PROJECT)\n"
] ]
}, },
{ {

141
notebooks/sensitivity_analysis.ipynb
View File

@ -39,27 +39,36 @@
}, },
{ {
"cell_type": "code", "cell_type": "code",
"execution_count": 2, "execution_count": 5,
"metadata": {}, "metadata": {},
"outputs": [ "outputs": [
{ {
"name": "stderr", "name": "stderr",
"output_type": "stream", "output_type": "stream",
"text": [ "text": [
"Failed to detect the name of this notebook, you can set it manually with the WANDB_NOTEBOOK_NAME environment variable to enable code saving.\n", "\u001b[34m\u001b[1mwandb\u001b[0m: Downloading large artifact btc-usdt-1m:latest, 3717.80MB. 12 files... \n",
"\u001b[34m\u001b[1mwandb\u001b[0m: 12 of 12 files downloaded. \n",
"Done. 0:0:5.3\n",
"\u001b[34m\u001b[1mwandb\u001b[0m: Downloading large artifact btc-usdt-5m:latest, 745.12MB. 12 files... \n", "\u001b[34m\u001b[1mwandb\u001b[0m: Downloading large artifact btc-usdt-5m:latest, 745.12MB. 12 files... \n",
"\u001b[34m\u001b[1mwandb\u001b[0m: 12 of 12 files downloaded. \n", "\u001b[34m\u001b[1mwandb\u001b[0m: 12 of 12 files downloaded. \n",
"Done. 0:0:1.1\n", "Done. 0:0:0.8\n",
"\u001b[34m\u001b[1mwandb\u001b[0m: Downloading large artifact btc-usdt-15m:latest, 248.65MB. 12 files... \n", "\u001b[34m\u001b[1mwandb\u001b[0m: Downloading large artifact btc-usdt-15m:latest, 248.65MB. 12 files... \n",
"\u001b[34m\u001b[1mwandb\u001b[0m: 12 of 12 files downloaded. \n", "\u001b[34m\u001b[1mwandb\u001b[0m: 12 of 12 files downloaded. \n",
"Done. 0:0:0.5\n", "Done. 0:0:0.5\n",
"\u001b[34m\u001b[1mwandb\u001b[0m: Downloading large artifact btc-usdt-30m:latest, 124.19MB. 12 files... \n", "\u001b[34m\u001b[1mwandb\u001b[0m: Downloading large artifact btc-usdt-30m:latest, 124.19MB. 12 files... \n",
"\u001b[34m\u001b[1mwandb\u001b[0m: 12 of 12 files downloaded. \n", "\u001b[34m\u001b[1mwandb\u001b[0m: 12 of 12 files downloaded. \n",
"Done. 0:0:0.4\n" "Done. 0:0:0.5\n"
] ]
} }
], ],
"source": [ "source": [
"data_windows_1min = get_data_windows(\n",
" 'wne-masters-thesis-testing',\n",
" 'btc-usdt-1m:latest',\n",
" min_window=0, \n",
" max_window=5\n",
")\n",
"\n",
"data_windows_5min = get_data_windows(\n", "data_windows_5min = get_data_windows(\n",
" 'wne-masters-thesis-testing',\n", " 'wne-masters-thesis-testing',\n",
" 'btc-usdt-5m:latest',\n", " 'btc-usdt-5m:latest',\n",
@ -84,10 +93,13 @@
}, },
{ {
"cell_type": "code", "cell_type": "code",
"execution_count": 3, "execution_count": 6,
"metadata": {}, "metadata": {},
"outputs": [], "outputs": [],
"source": [ "source": [
"with open('cache/1min-best-strategies-v1.pkl', 'rb') as inpt:\n",
" best_strategies_1min = pickle.load(inpt)\n",
"\n",
"with open('cache/5min-best-strategies-v2.pkl', 'rb') as inpt:\n", "with open('cache/5min-best-strategies-v2.pkl', 'rb') as inpt:\n",
" best_strategies_5min = pickle.load(inpt)\n", " best_strategies_5min = pickle.load(inpt)\n",
"\n", "\n",
@ -107,31 +119,34 @@
}, },
{ {
"cell_type": "code", "cell_type": "code",
"execution_count": 7, "execution_count": 10,
"metadata": {}, "metadata": {},
"outputs": [], "outputs": [],
"source": [ "source": [
"def results_plot(\n", "def results_plot(\n",
" buy_and_hold_concat,\n", " buy_and_hold_concat,\n",
" macd_30min_concat,\n", " # macd_30min_concat,\n",
" rsi_30min_concat,\n", " # rsi_30min_concat,\n",
" rsi_5_min_concat,\n", " # rsi_5_min_concat,\n",
" rmse_30min_model_concat,\n", " # rmse_30min_model_concat,\n",
" rmse_15min_model_concat,\n", " # rmse_15min_model_concat,\n",
" gmadl_30min_model_concat,\n", " gmadl_30min_model_concat,\n",
" gmadl_15min_model_concat,\n", " gmadl_15min_model_concat,\n",
" gmadl_5min_model_concat, width=850, height=500, notitle=False):\n", " gmadl_5min_model_concat, \n",
" gmadl_1min_model_concat, \n",
" width=850, height=500, notitle=False):\n",
"\n", "\n",
" fig = go.Figure([\n", " fig = go.Figure([\n",
" go.Scatter(y=buy_and_hold_concat['portfolio_value'], x=buy_and_hold_concat['time'], name=\"Buy and Hold\"),\n", " go.Scatter(y=buy_and_hold_concat['portfolio_value'], x=buy_and_hold_concat['time'], name=\"Buy and Hold\"),\n",
" go.Scatter(y=macd_30min_concat['portfolio_value'], x=macd_30min_concat['time'], name='MACD Strategy (30min)'),\n", " # go.Scatter(y=macd_30min_concat['portfolio_value'], x=macd_30min_concat['time'], name='MACD Strategy (30min)'),\n",
" go.Scatter(y=rsi_30min_concat['portfolio_value'], x=rsi_30min_concat['time'], name='RSI Strategy (30min)'),\n", " # go.Scatter(y=rsi_30min_concat['portfolio_value'], x=rsi_30min_concat['time'], name='RSI Strategy (30min)'),\n",
" go.Scatter(y=rsi_5_min_concat['portfolio_value'], x=rsi_5_min_concat['time'], name=\"RSI Strategy (5min)\"),\n", " # go.Scatter(y=rsi_5_min_concat['portfolio_value'], x=rsi_5_min_concat['time'], name=\"RSI Strategy (5min)\"),\n",
" go.Scatter(y=rmse_30min_model_concat['portfolio_value'], x=rmse_30min_model_concat['time'], name='RMSE Informer Strategy (30min)'),\n", " # go.Scatter(y=rmse_30min_model_concat['portfolio_value'], x=rmse_30min_model_concat['time'], name='RMSE Informer Strategy (30min)'),\n",
" go.Scatter(y=rmse_15min_model_concat['portfolio_value'], x=rmse_15min_model_concat['time'], name='RMSE Informer Strategy (15min)'),\n", " # go.Scatter(y=rmse_15min_model_concat['portfolio_value'], x=rmse_15min_model_concat['time'], name='RMSE Informer Strategy (15min)'),\n",
" go.Scatter(y=gmadl_30min_model_concat['portfolio_value'], x=gmadl_30min_model_concat['time'], name='GMADL Informer Strategy (30min)'),\n", " go.Scatter(y=gmadl_30min_model_concat['portfolio_value'], x=gmadl_30min_model_concat['time'], name='GMADL Informer Strategy (30min)'),\n",
" go.Scatter(y=gmadl_15min_model_concat['portfolio_value'], x=gmadl_15min_model_concat['time'], name='GMADL Informer Strategy (15min)'),\n", " go.Scatter(y=gmadl_15min_model_concat['portfolio_value'], x=gmadl_15min_model_concat['time'], name='GMADL Informer Strategy (15min)'),\n",
" go.Scatter(y=gmadl_5min_model_concat['portfolio_value'], x=gmadl_5min_model_concat['time'], name=\"GMADL Informer Strategy (5min)\")\n", " go.Scatter(y=gmadl_5min_model_concat['portfolio_value'], x=gmadl_5min_model_concat['time'], name=\"GMADL Informer Strategy (5min)\"),\n",
" go.Scatter(y=gmadl_1min_model_concat['portfolio_value'], x=gmadl_1min_model_concat['time'], name=\"GMADL Informer Strategy (1min)\")\n",
" ])\n", " ])\n",
" fig.update_layout(\n", " fig.update_layout(\n",
" # title={\n", " # title={\n",
@ -178,14 +193,16 @@
" \n", " \n",
"def results_table(\n", "def results_table(\n",
" buy_and_hold_concat,\n", " buy_and_hold_concat,\n",
" macd_30min_concat,\n", " # macd_30min_concat,\n",
" rsi_30min_concat,\n", " # rsi_30min_concat,\n",
" rsi_5_min_concat,\n", " # rsi_5_min_concat,\n",
" rmse_30min_model_concat,\n", " # rmse_30min_model_concat,\n",
" rmse_15min_model_concat,\n", " # rmse_15min_model_concat,\n",
" gmadl_30min_model_concat,\n", " gmadl_30min_model_concat,\n",
" gmadl_15min_model_concat,\n", " gmadl_15min_model_concat,\n",
" gmadl_5min_model_concat):\n", " gmadl_5min_model_concat,\n",
" gmadl_1min_model_concat\n",
" ):\n",
"\n", "\n",
" table_eval_windows = Texttable()\n", " table_eval_windows = Texttable()\n",
" table_eval_windows.set_deco(Texttable.HEADER)\n", " table_eval_windows.set_deco(Texttable.HEADER)\n",
@ -207,14 +224,15 @@
"\n", "\n",
" strategy_name_result = [\n", " strategy_name_result = [\n",
" ('Buy and Hold', buy_and_hold_concat),\n", " ('Buy and Hold', buy_and_hold_concat),\n",
" ('MACD Strategy (30min)', macd_30min_concat),\n", " # ('MACD Strategy (30min)', macd_30min_concat),\n",
" ('RSI Strategy (30min)', rsi_30min_concat),\n", " # ('RSI Strategy (30min)', rsi_30min_concat),\n",
" ('RSI Strategy (5min)', rsi_5_min_concat),\n", " # ('RSI Strategy (5min)', rsi_5_min_concat),\n",
" ('RMSE Informer (30min)', rmse_30min_model_concat),\n", " # ('RMSE Informer (30min)', rmse_30min_model_concat),\n",
" ('RMSE Informer (15min)', rmse_15min_model_concat),\n", " # ('RMSE Informer (15min)', rmse_15min_model_concat),\n",
" ('GMADL Informer (30min)', gmadl_30min_model_concat),\n", " ('GMADL Informer (30min)', gmadl_30min_model_concat),\n",
" ('GMADL Informer (15min)', gmadl_15min_model_concat),\n", " ('GMADL Informer (15min)', gmadl_15min_model_concat),\n",
" ('GMADL Informer (5min)', gmadl_5min_model_concat),\n", " ('GMADL Informer (5min)', gmadl_5min_model_concat),\n",
" ('GMADL Informer (1min)', gmadl_1min_model_concat),\n",
" ]\n", " ]\n",
" for strategy_name, result in strategy_name_result:\n", " for strategy_name, result in strategy_name_result:\n",
" table_eval_windows.add_row([\n", " table_eval_windows.add_row([\n",
@ -234,58 +252,50 @@
}, },
{ {
"cell_type": "code", "cell_type": "code",
"execution_count": 9, "execution_count": 1,
"metadata": {}, "metadata": {},
"outputs": [ "outputs": [
{ {
"name": "stdout", "ename": "NameError",
"output_type": "stream", "evalue": "name 'pd' is not defined",
"text": [ "output_type": "error",
"\\begin{table}\n", "traceback": [
"\t\\begin{center}\n", "\u001b[0;31m---------------------------------------------------------------------------\u001b[0m",
"\t\t\\begin{tabular}{lccccccccc}\n", "\u001b[0;31mNameError\u001b[0m Traceback (most recent call last)",
"\t\t\t\\textbf{Strategy} & \\textbf{VAL} & \\textbf{ARC} & \\textbf{ASD} & \\textbf{IR*} & \\textbf{MD} & \\textbf{IR**} & \\textbf{N} & \\textbf{LONG} & \\textbf{SHORT} \\\\\n", "Cell \u001b[0;32mIn[1], line 1\u001b[0m\n\u001b[0;32m----> 1\u001b[0m test_data_1min \u001b[38;5;241m=\u001b[39m \u001b[43mpd\u001b[49m\u001b[38;5;241m.\u001b[39mconcat([data_windows_1min[\u001b[38;5;241m0\u001b[39m][\u001b[38;5;241m0\u001b[39m][\u001b[38;5;241m-\u001b[39mPADDING:]] \u001b[38;5;241m+\u001b[39m [data_window[\u001b[38;5;241m1\u001b[39m] \u001b[38;5;28;01mfor\u001b[39;00m data_window \u001b[38;5;129;01min\u001b[39;00m data_windows_1min])\n\u001b[1;32m 2\u001b[0m test_data_5min \u001b[38;5;241m=\u001b[39m pd\u001b[38;5;241m.\u001b[39mconcat([data_windows_5min[\u001b[38;5;241m0\u001b[39m][\u001b[38;5;241m0\u001b[39m][\u001b[38;5;241m-\u001b[39mPADDING:]] \u001b[38;5;241m+\u001b[39m [data_window[\u001b[38;5;241m1\u001b[39m] \u001b[38;5;28;01mfor\u001b[39;00m data_window \u001b[38;5;129;01min\u001b[39;00m data_windows_5min])\n\u001b[1;32m 3\u001b[0m test_data_15min \u001b[38;5;241m=\u001b[39m pd\u001b[38;5;241m.\u001b[39mconcat([data_windows_15min[\u001b[38;5;241m0\u001b[39m][\u001b[38;5;241m0\u001b[39m][\u001b[38;5;241m-\u001b[39mPADDING:]] \u001b[38;5;241m+\u001b[39m [data_window[\u001b[38;5;241m1\u001b[39m] \u001b[38;5;28;01mfor\u001b[39;00m data_window \u001b[38;5;129;01min\u001b[39;00m data_windows_15min])\n",
"\t\t\t\\hline\n", "\u001b[0;31mNameError\u001b[0m: name 'pd' is not defined"
"\t\t\tBuy and Hold & 1.441 & 0.131 & 0.577 & 0.228 & 0.773 & 0.039 & 2 & 100.00\\% & 0.00\\% \\\\\n",
"\t\t\tMACD Strategy (30min) & 1.952 & 0.254 & 0.524 & 0.485 & 0.592 & 0.207 & 327 & 52.30\\% & 28.30\\% \\\\\n",
"\t\t\tRSI Strategy (30min) & 4.542 & 0.668 & 0.462 & 1.444 & 0.399 & 2.415 & 377 & 30.79\\% & 28.03\\% \\\\\n",
"\t\t\tRSI Strategy (5min) & 3.341 & 0.503 & 0.504 & 0.999 & 0.300 & 1.676 & 846 & 28.29\\% & 33.47\\% \\\\\n",
"\t\t\tRMSE Informer (30min) & 2.727 & 0.404 & 0.505 & 0.800 & 0.518 & 0.624 & 34 & 64.40\\% & 24.67\\% \\\\\n",
"\t\t\tRMSE Informer (15min) & 1.509 & 0.149 & 0.349 & 0.428 & 0.455 & 0.140 & 16 & 15.24\\% & 27.60\\% \\\\\n",
"\t\t\tGMADL Informer (30min) & 2.263 & 0.318 & 0.367 & 0.866 & 0.533 & 0.516 & 811 & 35.51\\% & 19.59\\% \\\\\n",
"\t\t\tGMADL Informer (15min) & 3.296 & 0.496 & 0.527 & 0.942 & 0.474 & 0.987 & 362 & 49.37\\% & 37.72\\% \\\\\n",
"\t\t\tGMADL Informer (5min) & 9.747 & 1.159 & 0.544 & 2.129 & 0.327 & 7.552 & 846 & 44.80\\% & 41.51\\% \\\\\n",
"\t\t\\end{tabular}\n",
"\t\\end{center}\n",
"\\end{table}\n"
] ]
} }
], ],
"source": [ "source": [
"test_data_1min = pd.concat([data_windows_1min[0][0][-PADDING:]] + [data_window[1] for data_window in data_windows_1min])\n",
"test_data_5min = pd.concat([data_windows_5min[0][0][-PADDING:]] + [data_window[1] for data_window in data_windows_5min])\n", "test_data_5min = pd.concat([data_windows_5min[0][0][-PADDING:]] + [data_window[1] for data_window in data_windows_5min])\n",
"test_data_15min = pd.concat([data_windows_15min[0][0][-PADDING:]] + [data_window[1] for data_window in data_windows_15min])\n", "test_data_15min = pd.concat([data_windows_15min[0][0][-PADDING:]] + [data_window[1] for data_window in data_windows_15min])\n",
"test_data_30min = pd.concat([data_windows_30min[0][0][-PADDING:]] + [data_window[1] for data_window in data_windows_30min])\n", "test_data_30min = pd.concat([data_windows_30min[0][0][-PADDING:]] + [data_window[1] for data_window in data_windows_30min])\n",
"\n", "\n",
"buy_and_hold_concat = evaluate_strategy(test_data_5min, BuyAndHoldStrategy(), padding=PADDING, interval='5min')\n", "buy_and_hold_concat = evaluate_strategy(test_data_5min, BuyAndHoldStrategy(), padding=PADDING, interval='5min')\n",
"macd_30min_concat = evaluate_strategy(test_data_30min, ConcatenatedStrategies(len(data_windows_30min[0][1]), [s[0] for s in best_strategies_30min['macd_strategies']], padding=PADDING), padding=PADDING, interval='30min')\n", "# macd_30min_concat = evaluate_strategy(test_data_30min, ConcatenatedStrategies(len(data_windows_30min[0][1]), [s[0] for s in best_strategies_30min['macd_strategies']], padding=PADDING), padding=PADDING, interval='30min')\n",
"rsi_30min_concat = evaluate_strategy(test_data_30min, ConcatenatedStrategies(len(data_windows_30min[0][1]), [s[0] for s in best_strategies_30min['rsi_strategies']], padding=PADDING), padding=PADDING, interval='30min')\n", "# rsi_30min_concat = evaluate_strategy(test_data_30min, ConcatenatedStrategies(len(data_windows_30min[0][1]), [s[0] for s in best_strategies_30min['rsi_strategies']], padding=PADDING), padding=PADDING, interval='30min')\n",
"rsi_5_min_concat = evaluate_strategy(test_data_5min, ConcatenatedStrategies(len(data_windows_5min[0][1]), [s[0] for s in best_strategies_5min['rsi_strategies']], padding=PADDING), padding=PADDING, interval='5min')\n", "# rsi_5_min_concat = evaluate_strategy(test_data_5min, ConcatenatedStrategies(len(data_windows_5min[0][1]), [s[0] for s in best_strategies_5min['rsi_strategies']], padding=PADDING), padding=PADDING, interval='5min')\n",
"rmse_30min_model_concat = evaluate_strategy(test_data_30min, ConcatenatedStrategies(len(data_windows_30min[0][1]), [s[0] for s in best_strategies_30min['rmse_model']], padding=PADDING), padding=PADDING, interval='30min')\n", "# rmse_30min_model_concat = evaluate_strategy(test_data_30min, ConcatenatedStrategies(len(data_windows_30min[0][1]), [s[0] for s in best_strategies_30min['rmse_model']], padding=PADDING), padding=PADDING, interval='30min')\n",
"rmse_15min_model_concat = evaluate_strategy(test_data_15min, ConcatenatedStrategies(len(data_windows_15min[0][1]), [s[0] for s in best_strategies_15min['rmse_model']], padding=PADDING), padding=PADDING, interval='15min')\n", "# rmse_15min_model_concat = evaluate_strategy(test_data_15min, ConcatenatedStrategies(len(data_windows_15min[0][1]), [s[0] for s in best_strategies_15min['rmse_model']], padding=PADDING), padding=PADDING, interval='15min')\n",
"gmadl_30min_model_concat = evaluate_strategy(test_data_30min, ConcatenatedStrategies(len(data_windows_30min[0][1]), [s[0] for s in best_strategies_30min['gmadl_model']], padding=PADDING), padding=PADDING, interval='30min')\n", "gmadl_30min_model_concat = evaluate_strategy(test_data_30min, ConcatenatedStrategies(len(data_windows_30min[0][1]), [s[0] for s in best_strategies_30min['gmadl_model']], padding=PADDING), padding=PADDING, interval='30min')\n",
"gmadl_15min_model_concat = evaluate_strategy(test_data_15min, ConcatenatedStrategies(len(data_windows_15min[0][1]), [s[0] for s in best_strategies_15min['gmadl_model']], padding=PADDING), padding=PADDING, interval='15min')\n", "gmadl_15min_model_concat = evaluate_strategy(test_data_15min, ConcatenatedStrategies(len(data_windows_15min[0][1]), [s[0] for s in best_strategies_15min['gmadl_model']], padding=PADDING), padding=PADDING, interval='15min')\n",
"gmadl_1min_model_concat = evaluate_strategy(test_data_1min, ConcatenatedStrategies(len(data_windows_1min[0][1]), [s[0] for s in best_strategies_1min['gmadl_model']], padding=PADDING), padding=PADDING, interval='min')\n",
"gmadl_5min_model_concat = evaluate_strategy(test_data_5min, ConcatenatedStrategies(len(data_windows_5min[0][1]), [s[0] for s in best_strategies_5min['gmadl_model']], padding=PADDING), padding=PADDING, interval='5min')\n", "gmadl_5min_model_concat = evaluate_strategy(test_data_5min, ConcatenatedStrategies(len(data_windows_5min[0][1]), [s[0] for s in best_strategies_5min['gmadl_model']], padding=PADDING), padding=PADDING, interval='5min')\n",
"\n", "\n",
"results_table(\n", "# results_table(\n",
" buy_and_hold_concat,\n", "# buy_and_hold_concat,\n",
" macd_30min_concat,\n", " # macd_30min_concat,\n",
" rsi_30min_concat,\n", " # rsi_30min_concat,\n",
" rsi_5_min_concat,\n", " # rsi_5_min_concat,\n",
" rmse_30min_model_concat,\n", " # rmse_30min_model_concat,\n",
" rmse_15min_model_concat,\n", " # rmse_15min_model_concat,\n",
" gmadl_30min_model_concat,\n", " # gmadl_30min_model_concat,\n",
" gmadl_15min_model_concat,\n", " # gmadl_15min_model_concat,\n",
" gmadl_5min_model_concat)\n", " # gmadl_5min_model_concat,\n",
" # gmadl_1min_model_concat\n",
" # )\n",
"\n", "\n",
"# results_plot(\n", "# results_plot(\n",
"# buy_and_hold_concat,\n", "# buy_and_hold_concat,\n",
@ -297,6 +307,7 @@
" # gmadl_30min_model_concat,\n", " # gmadl_30min_model_concat,\n",
" # gmadl_15min_model_concat,\n", " # gmadl_15min_model_concat,\n",
" # gmadl_5min_model_concat, \n", " # gmadl_5min_model_concat, \n",
" # gmadl_1min_model_concat, \n",
" # width=1200, notitle=True)" " # width=1200, notitle=True)"
] ]
}, },