the plot function should be revised. now that the result has been obtained, plotting fig is easy.

from finrl.

"""
Script for comparing results with DJIA.

This script reads data from df_result_ensemble and df_dji, merges them, and visualizes the comparison.
"""

import necessary packages

import pandas as pd
import os
import matplotlib.pyplot as plt
from finrl.config import DATA_SAVE_DIR

Load df_result_ensemble data from the CSV file

df_result_ensemble = pd.read_csv(
os.path.join(DATA_SAVE_DIR, "df_result_ensemble.csv"), index_col="date"
)

load df_dji data from csv

df_dji = pd.read_csv(os.path.join(DATA_SAVE_DIR, "df_dji.csv"), index_col="date")

Merge DataFrames on index

result = pd.merge(df_result_ensemble, df_dji, left_index=True, right_index=True)

Save to CSV

result.to_csv(os.path.join(DATA_SAVE_DIR, "result.csv"))

Rename columns for clarity

result.columns = ["ensemble", "djia"]

Plotting the result

Set the backend to a suitable interactive backend

plt.switch_backend("TkAgg") # Use 'TkAgg' backend for Tkinter-based interactive display
plt.rcParams["figure.figsize"] = (15, 5)
plt.figure()
result.plot()
plt.title("Ensemble and DJIA Comparison")
plt.xlabel("Date")
plt.ylabel("Values")

Display the plot interactively

plt.show()

from finrl.

@VishnuRaghavan
can your code successfully plot the figure? if yes, pls submit a PR

from finrl.

@zhumingpassional I'm running on my local machine I have split the colab file into multiple files the above code is for comparison with DJIA. Also I'm training it on my CUDA NVIDIA GPU. Making a PR would not be the best solution as the file architectures are now different.

And yes for me plotting is working fine with this setup.

from finrl.