01/11/2024, 15:41 Untitled11.
ipynb - Colab
!pip install meteostat pandas
Collecting meteostat
Downloading [Link] (4.6 kB)
Requirement already satisfied: pandas in /usr/local/lib/python3.10/dist-packag
Requirement already satisfied: pytz in /usr/local/lib/python3.10/dist-packages
Requirement already satisfied: numpy in /usr/local/lib/python3.10/dist-package
Requirement already satisfied: python-dateutil>=2.8.2 in /usr/local/lib/python
Requirement already satisfied: tzdata>=2022.7 in /usr/local/lib/python3.10/dis
Requirement already satisfied: six>=1.5 in /usr/local/lib/python3.10/dist-pack
Downloading [Link] (31 kB)
Installing collected packages: meteostat
Successfully installed meteostat-1.6.8
from datetime import datetime, timedelta
import pandas as pd
from meteostat import Point, Daily
mesra = Point(23.4294, 85.4124, 350)
today = [Link]()
start_date = today - timedelta(days=60)
end_date = today
data = Daily(mesra, start_date, end_date)
data = [Link]()
if 'tmean' in [Link]:
temperatures = data[['tmean']]
else:
print("Warning: 'tmean' column not found in the data. Using 'tavg' instead.")
temperatures = data[['tavg']]
Warning: 'tmean' column not found in the data. Using 'tavg' instead.
temperatures.to_csv('[Link]', index=True)
[Link] 1/8
01/11/2024, 15:41 [Link] - Colab
import pandas as pd
temp_df = pd.read_csv('[Link]', index_col=0)
temp_df
Show hidden output
Next
steps:
Generate code
with
temp_df
toggle_off View recommended
plots
New interactive
sheet
import pandas as pd
from [Link] import MinMaxScaler
temp_df = pd.read_csv('[Link]', index_col=0)
temp_col_name = temp_df.columns[0]
scaler = MinMaxScaler()
temp_values = temp_df[temp_col_name].[Link](-1, 1)
normalized_temp = scaler.fit_transform(temp_values)
temp_df['Normalized_Temperature'] = normalized_temp
temp_df
[Link] 2/8
01/11/2024, 15:41 [Link] - Colab
tavg Normalized_Temperature
time
2024-09-03 27.9 0.815385
2024-09-04 28.2 0.861538
2024-09-05 27.9 0.815385
2024-09-06 28.1 0.846154
2024-09-07 29.1 1.000000
2024-09-08 28.6 0.923077
2024-09-09 27.2 0.707692
2024-09-10 27.0 0.676923
2024-09-11 27.8 0.800000
2024-09-12 28.3 0.876923
2024-09-13 27.8 0.800000
2024-09-14 25.3 0.415385
2024-09-15 25.0 0.369231
2024-09-16 24.7 0.323077
2024-09-17 25.5 0.446154
2024-09-18 26.4 0.584615
2024-09-19 28.1 0.846154
2024-09-20 28.7 0.938462
2024-09-21 29.0 0.984615
2024-09-22 28.7 0.938462
2024-09-23 28.1 0.846154
2024-09-24 27.1 0.692308
2024-09-25 25.2 0.400000
2024-09-26 26.2 0.553846
2024-09-27 26.1 0.538462
2024-09-28 28.0 0.830769
2024-09-29 28.2 0.861538
2024-09-30 28.5 0.907692
2024-10-01 28.5 0.907692
2024-10-02 27.8 0.800000
2024-10-03 27.8 0.800000
[Link] 3/8
01/11/2024, 15:41 [Link] - Colab
2024-10-04 27.7 0.784615
2024-10-05 27.6 0.769231
2024-10-06 26.7 0.630769
2024-10-07 27.1 0.692308
2024-10-08 26.5 0.600000
2024-10-09 26.4 0.584615
2024-10-10 26.3 0.569231
2024-10-11 26.9 0.661538
2024-10-12 26.9 0.661538
2024-10-13 26.4 0.584615
2024-10-14 25.6 0.461538
2024-10-15 25.3 0.415385
2024-10-16 25.8 0.492308
2024-10-17 26.1 0.538462
2024-10-18 26.0 0.523077
2024-10-19 26.4 0.584615
2024-10-20 25.6 0.461538
2024-10-21 25.4 0.430769
2024-10-22 24.7 0.323077
2024-10-23 24.9 0.353846
2024-10-24 23.4 0.123077
2024-10-25 22.6 0.000000
2024-10-26 23.1 0.076923
2024-10-27 25.2 0.400000
2024-10-28 26.0 0.523077
2024-10-29 26.0 0.523077
2024-10-30 26.2 0.553846
2024-10-31 25.4 0.430769
2024-11-01 24.0 0.215385
[Link] 4/8
01/11/2024, 15:41 [Link] - Colab
Next
steps:
Generate code
with
temp_df
toggle_off View recommended
plots
New interactive
sheet
def create_sequences(data, n_input_days):
X, y = [], []
for i in range(len(data) - n_input_days):
[Link](data[i:i + n_input_days])
[Link](data[i + n_input_days])
return X, y
normalized_temps = [0.1, 0.2, 0.3, 0.4, 0.5, 0.6]
n_input_days = 3
X, y = create_sequences(normalized_temps, n_input_days)
print("X:", X)
print("y:", y)
X: [[0.1, 0.2, 0.3], [0.2, 0.3, 0.4], [0.3, 0.4, 0.5]]
y: [0.4, 0.5, 0.6]
import numpy as np
import pandas as pd
from [Link] import MinMaxScaler
from [Link] import Sequential
from [Link] import LSTM, Dense
normalized_temps = temp_df['Normalized_Temperature'].values
n_input_days = 3
def create_sequences(data, n_input_days):
X, y = [], []
for i in range(len(data) - n_input_days):
[Link](data[i:i + n_input_days])
[Link](data[i + n_input_days])
return [Link](X), [Link](y)
X, y = create_sequences(normalized_temps, n_input_days)
model = Sequential()
[Link](LSTM(units=50, activation='relu', input_shape=([Link][1], 1)))
[Link](Dense(units=1))
[Link](loss='mse', optimizer='adam')
[Link](X, y, epochs=50, batch_size=32)
[Link] 5/8
01/11/2024, 15:41 [Link] - Colab
[Link] 6/8
01/11/2024, 15:41 [Link] - Colab
<[Link] at 0x7802688087c0>
last_n_days_temps = normalized_temps[-n_input_days:]
input_sequence = [Link]([last_n_days_temps])
prediction = [Link](input_sequence)
actual_prediction = scaler.inverse_transform(prediction)
print("Predicted temperature for the next day:", actual_prediction[0][0])
1/1 ━━━━━━━━━━━━━━━━━━━━ 1s 533ms/step
Predicted temperature for the next day: 25.390377
import numpy as np
from tensorflow import keras
from [Link] import Sequential
from [Link] import LSTM, Dense
import numpy as np
from tensorflow import keras
from [Link] import Sequential
from [Link] import SimpleRNN, Dense
sequence = [10, 20, 30, 40, 50, 60, 70, 80, 90]
def create_sequences(data, seq_length):
X, y = [], []
for i in range(len(data) - seq_length - 1):
[Link](data[i:(i + seq_length)])
[Link](data[i + seq_length])
return [Link](X), [Link](y)
seq_length = 3
X, y = create_sequences(sequence, seq_length)
X = [Link](([Link][0], [Link][1], 1))
model = Sequential()
[Link](SimpleRNN(units=50, activation='relu', input_shape=(seq_length, 1)))
[Link](Dense(units=1))
[Link](loss='mse', optimizer='adam')
[Link] 7/8
01/11/2024, 15:41 [Link] - Colab
[Link](X, y, epochs=200, verbose=0)
input_seq = [Link](sequence[-seq_length:]).reshape((1, seq_length, 1))
predicted_number = [Link](input_seq)[0][0]
print("Predicted next number:", predicted_number)
UserWarning: Do not pass an `input_shape`/`input_dim` argument to a layer. Whe
1/1 ━━━━━━━━━━━━━━━━━━━━ 0s 256ms/step
Predicted next number: 110.82568
import numpy as np
from tensorflow import keras
from [Link] import Sequential
from [Link] import LSTM, Dense
text = "This is a sample text sequence."
chars = sorted(list(set(text)))
char_to_index = {char: index for index, char in enumerate(chars)}
index_to_char = {index: char for index, char in enumerate(chars)}
seq_length = 3
X, y = [], []
for i in range(len(text) - seq_length - 1):
[Link]([char_to_index[char] for char in text[i:i + seq_length]])
[Link](char_to_index[text[i + seq_length]])
X = [Link](X)
y = [Link](y)
X = [Link](([Link][0], [Link][1], 1))
X = X / float(len(chars))
model = Sequential()
[Link](LSTM(units=50, activation='relu', input_shape=(seq_length, 1)))
[Link](Dense(units=len(chars), activation='softmax'))
[Link](loss='sparse_categorical_crossentropy', optimizer='adam')
[Link](X, y, epochs=100, verbose=0)
[Link] 8/8
