Not sure what the error means. Its in the same X_train, y_train format as accuracy, so not sure why those 3 wouldnt work.
Here is my code:
cv_scores = []
cv_std = []
# Logistic Regression
logreg = make_pipeline(RobustScaler(), LogisticRegression())
logreg.fit(X_train, y_train)
preds = logreg.predict(X_train)
acc_log = round(logreg.score(X_train, y_train) * 100, 2)
cv_scores.append(acc_log)
print(accuracy_score(y_train, preds))
print('Precision is: ', precision_score(y_train, preds))
print('Recall is: ', recall_score(y_train, preds))
print('F1 is: ', f1_score(y_train, preds))
And this is my output:
Accuracy is: 0.8372615039281706
/usr/local/lib/python3.8/dist-packages/sklearn/metrics/_classification.py in precision_score(y_true, y_pred, labels, pos_label, average, sample_weight, zero_division)
1952 array([0.5, 1. , 1. ])
1953 """
-> 1954 p, _, _, _ = precision_recall_fscore_support(
1955 y_true,
1956 y_pred,
1 frames
/usr/local/lib/python3.8/dist-packages/sklearn/metrics/_classification.py in _check_set_wise_labels(y_true, y_pred, average, labels, pos_label)
1380 if pos_label not in present_labels:
1381 if len(present_labels) >= 2:
-> 1382 raise ValueError(
1383 f"pos_label={pos_label} is not a valid label. It "
1384 f"should be one of {present_labels}"
ValueError: pos_label=1 is not a valid label. It should be one of ['0', '1']
X_train here:
{'Pclass': {0: 1.1943176378757767, 1: 0.0, 2: 1.1943176378757767, 3: 0.0, 4: 1.1943176378757767}, 'Age': {0: 4.003419248109409, 1: 4.882972919330174, 2: 4.2631607783924865, 3: 4.745132396064701, 4: 4.745132396064701}, 'SibSp': {0: 0.7304631471189666, 1: 0.7304631471189666, 2: 0.0, 3: 0.7304631471189666, 4: 0.0}, 'Parch': {0: 0.0, 1: 0.0, 2: 0.0, 3: 0.0, 4: 0.0}, 'Fare': {0: 0.0, 1: 1.5409627556327752, 2: 0.7304631471189666, 3: 1.5409627556327752, 4: 0.7304631471189666}, 'FamilySize': {0: 1.1943176378757767, 1: 1.1943176378757767, 2: 0.7304631471189666, 3: 1.1943176378757767, 4: 0.7304631471189666}, 'Age*Class': {0: 5.133567292834429, 1: 0.0, 2: 5.42678077191518, 3: 0.0, 4: 5.968980585391403}, 'Sex_female': {0: 0.0, 1: 1.0, 2: 1.0, 3: 1.0, 4: 0.0}, 'Sex_male': {0: 1.0, 1: 0.0, 2: 0.0, 3: 0.0, 4: 1.0}, 'Embarked_C': {0: 0.0, 1: 1.0, 2: 0.0, 3: 0.0, 4: 0.0}, 'Embarked_Q': {0: 0.0, 1: 0.0, 2: 0.0, 3: 0.0, 4: 0.0}, 'Embarked_S': {0: 1.0, 1: 0.0, 2: 1.0, 3: 1.0, 4: 1.0}, 'Cabin_Letter_A': {0: 0.0, 1: 0.0, 2: 0.0, 3: 0.0, 4: 0.0}, 'Cabin_Letter_B': {0: 0.0, 1: 0.0, 2: 0.0, 3: 0.0, 4: 0.0}, 'Cabin_Letter_C': {0: 1.0, 1: 1.0, 2: 1.0, 3: 1.0, 4: 1.0}, 'Cabin_Letter_D': {0: 0.0, 1: 0.0, 2: 0.0, 3: 0.0, 4: 0.0}, 'Cabin_Letter_E': {0: 0.0, 1: 0.0, 2: 0.0, 3: 0.0, 4: 0.0}, 'Cabin_Letter_F': {0: 0.0, 1: 0.0, 2: 0.0, 3: 0.0, 4: 0.0}, 'Cabin_Letter_G': {0: 0.0, 1: 0.0, 2: 0.0, 3: 0.0, 4: 0.0}, 'Cabin_Letter_T': {0: 0.0, 1: 0.0, 2: 0.0, 3: 0.0, 4: 0.0}, 'Ticket_Letter_A': {0: 1.0, 1: 0.0, 2: 0.0, 3: 0.0, 4: 0.0}, 'Ticket_Letter_C': {0: 0.0, 1: 0.0, 2: 0.0, 3: 0.0, 4: 0.0}, 'Ticket_Letter_CA': {0: 0.0, 1: 0.0, 2: 0.0, 3: 0.0, 4: 0.0}, 'Ticket_Letter_F': {0: 0.0, 1: 0.0, 2: 0.0, 3: 0.0, 4: 0.0}, 'Ticket_Letter_Fa': {0: 0.0, 1: 0.0, 2: 0.0, 3: 0.0, 4: 0.0}, 'Ticket_Letter_LINE': {0: 0.0, 1: 0.0, 2: 0.0, 3: 0.0, 4: 0.0}, 'Ticket_Letter_P': {0: 0.0, 1: 0.0, 2: 0.0, 3: 0.0, 4: 0.0}, 'Ticket_Letter_PC': {0: 0.0, 1: 1.0, 2: 0.0, 3: 1.0, 4: 1.0}, 'Ticket_Letter_PP': {0: 0.0, 1: 0.0, 2: 0.0, 3: 0.0, 4: 0.0}, 'Ticket_Letter_S': {0: 0.0, 1: 0.0, 2: 0.0, 3: 0.0, 4: 0.0}, 'Ticket_Letter_SC': {0: 0.0, 1: 0.0, 2: 0.0, 3: 0.0, 4: 0.0}, 'Ticket_Letter_SCO': {0: 0.0, 1: 0.0, 2: 0.0, 3: 0.0, 4: 0.0}, 'Ticket_Letter_SO': {0: 0.0, 1: 0.0, 2: 0.0, 3: 0.0, 4: 0.0}, 'Ticket_Letter_SOTON': {0: 0.0, 1: 0.0, 2: 0.0, 3: 0.0, 4: 0.0}, 'Ticket_Letter_STON': {0: 0.0, 1: 0.0, 2: 1.0, 3: 0.0, 4: 0.0}, 'Ticket_Letter_SW': {0: 0.0, 1: 0.0, 2: 0.0, 3: 0.0, 4: 0.0}, 'Ticket_Letter_W': {0: 0.0, 1: 0.0, 2: 0.0, 3: 0.0, 4: 0.0}, 'Ticket_Letter_WE': {0: 0.0, 1: 0.0, 2: 0.0, 3: 0.0, 4: 0.0}, 'Title_Master': {0: 0.0, 1: 0.0, 2: 0.0, 3: 0.0, 4: 0.0}, 'Title_Miss': {0: 0.0, 1: 0.0, 2: 1.0, 3: 0.0, 4: 0.0}, 'Title_Mr': {0: 1.0, 1: 0.0, 2: 0.0, 3: 0.0, 4: 1.0}, 'Title_Mrs': {0: 0.0, 1: 1.0, 2: 0.0, 3: 1.0, 4: 0.0}, 'Title_Rare': {0: 0.0, 1: 0.0, 2: 0.0, 3: 0.0, 4: 0.0}, 'IsAlone_0': {0: 1.0, 1: 1.0, 2: 0.0, 3: 1.0, 4: 0.0}, 'IsAlone_1': {0: 0.0, 1: 0.0, 2: 1.0, 3: 0.0, 4: 1.0}, 'Ticket_Letter_AQ': {0: 0, 1: 0, 2: 0, 3: 0, 4: 0}, 'Ticket_Letter_LP': {0: 0, 1: 0, 2: 0, 3: 0, 4: 0}}
y_train here:
{0: '0', 1: '1', 2: '1', 3: '1', 4: '0'}
Related
import numpy as np
from matplotlib import pyplot as plt
data = np.array([[0.8, 2.4, 2.5, 3.9, 0.0, 4.0, 0.0],
[2.4, 0.0, 4.0, 1.0, 2.7, 0.0, 0.0],
[1.1, 2.4, 0.8, 4.3, 1.9, 4.4, 0.0],
[0.6, 0.0, 0.3, 0.0, 3.1, 0.0, 0.0],
[0.7, 1.7, 0.6, 2.6, 2.2, 6.2, 0.0],
[1.3, 1.2, 0.0, 0.0, 0.0, 3.2, 5.1],
[0.1, 2.0, 0.0, 1.4, 0.0, 1.9, 6.3]])
plt.figure(figsize=(6, 4))
im = plt.imshow(data, cmap="YlGn")
linewidth = 2
for axis in ['top', 'bottom', 'left', 'right']:
plt.gca().spines[axis].set_linewidth(linewidth)
plt.gca().set_xticks(np.arange(data.shape[1] + 1) - .5, minor=True)
plt.gca().set_yticks(np.arange(data.shape[0] + 1) - .5, minor=True)
plt.gca().grid(which="minor", color="black", linewidth=linewidth)
plt.gca().tick_params(which="minor", bottom=False, left=False)
plt.tight_layout()
plt.gca().set_xticks(ticks=[])
plt.gca().set_yticks(ticks=[])
plt.savefig("test.pdf",
bbox_inches="tight",
transparent="True",
pad_inches=1.0/72.0 * linewidth / 2.0)
This code will output the following pdf, but you can see that there are white borders on the left and bottom, so the pdf is not centered after being inserted into LaTex. How to solve this problem?
plt result:
import numpy as np
from matplotlib import pyplot as plt
data = np.array([[0.8, 2.4, 2.5, 3.9, 0.0, 4.0, 0.0],
[2.4, 0.0, 4.0, 1.0, 2.7, 0.0, 0.0],
[1.1, 2.4, 0.8, 4.3, 1.9, 4.4, 0.0],
[0.6, 0.0, 0.3, 0.0, 3.1, 0.0, 0.0],
[0.7, 1.7, 0.6, 2.6, 2.2, 6.2, 0.0],
[1.3, 1.2, 0.0, 0.0, 0.0, 3.2, 5.1],
[0.1, 2.0, 0.0, 1.4, 0.0, 1.9, 6.3]])
plt.figure(figsize=(6, 4))
im = plt.imshow(data, cmap="YlGn")
linewidth = 2
for axis in ['top', 'bottom', 'left', 'right']:
plt.gca().spines[axis].set_linewidth(linewidth)
plt.gca().set_xticks(np.arange(data.shape[1] + 1) - .5, minor=True)
plt.gca().set_yticks(np.arange(data.shape[0] + 1) - .5, minor=True)
plt.gca().grid(which="minor", color="black", linewidth=linewidth)
plt.gca().tick_params(which="minor", bottom=False, left=False)
plt.tight_layout()
plt.gca().set_xticks(ticks=[])
plt.gca().set_yticks(ticks=[])
plt.gca().tick_params(axis="both",
which="major",
left=False,
bottom=False,
labelleft=False,
labelbottom=False)
plt.savefig("test.pdf",
bbox_inches="tight",
transparent="True",
pad_inches=1.0 / 72.0 * linewidth / 2.0)
It was an issue with ticks, solved now.
I'm trying to train LSTM model in Keras using data of variable timestep, for example, the data looks like:
<tf.RaggedTensor [[[0.0, 0.0, 0.0, 0.0, 0.0, 1.0, 1.0, 0.0, 1.0, 0.0, 0.0, 0.0, 0.0],
[0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 1.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0]],
[[1.0, 0.0, 0.0, 0.0, 0.0, 1.0, 1.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0],
[1.0, 0.0, 0.0, 0.0, 0.0, 1.0, 1.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0]],
[[0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 1.0, 0.0, 0.0]], ...,
[[0.0, 1.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0]],
[[1.0, 1.0, 1.0, 0.0, 0.0, 0.0, 1.0, 1.0, 0.0, 0.0, 1.0, 0.0, 0.0],
[1.0, 1.0, 0.0, 0.0, 0.0, 0.0, 1.0, 1.0, 0.0, 0.0, 1.0, 0.0, 0.0],
[1.0, 0.0, 0.0, 0.0, 0.0, 0.0, 1.0, 1.0, 0.0, 0.0, 1.0, 0.0, 0.0],
[1.0, 0.0, 0.0, 0.0, 0.0, 0.0, 1.0, 1.0, 0.0, 0.0, 1.0, 0.0, 0.0],
[1.0, 0.0, 0.0, 0.0, 0.0, 0.0, 1.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0]],
[[1.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0],
[1.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0],
[1.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0],
[1.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0],
[0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0]]]>
and its corresponding label:
<tf.RaggedTensor [[6, 6], [7, 7], [8], ..., [6], [11, 11, 11, 11, 11], [24, 24, 24, 24, 24]]>
Each input data have 13 features, so for each time step, the model receives a 1 x 13 vector. I wonder if it is possible to do so? I don't mind doing this on pytorch either.
I try to align them with no reshape layer.
However, my input for each time step in the LSTM layer is a vector of dimension 13. And each sample has variable-length of these vectors, which means the time step is not constant for each sample. Can you show me a code example of how to train such model? –
TurquoiseJ
First of all, the concept of windows length and time steps is they take the same amount of the input with a higher number of length and time.
We assume the input to extract features can be divide by multiple times of windows travels along with axis, please see the attached for idea.
[Codes]:
batched_features = tf.constant( [ [ 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0 ], [ 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0 ], ], shape=( 2, 1, 13 ) )
batched_labels = tf.constant( [[ 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0 ], [ 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0 ]], shape=( 2, 13 ) )
dataset = tf.data.Dataset.from_tensor_slices((batched_features, batched_labels))
dataset = dataset.batch(10)
batched_features = dataset
[Sample]:
Model: "sequential"
_________________________________________________________________
Layer (type) Output Shape Param #
=================================================================
bidirectional (Bidirectiona (None, 1, 64) 11776
l)
bidirectional_1 (Bidirectio (None, 64) 24832
nal)
dense (Dense) (None, 13) 845
=================================================================
Total params: 37,453
Trainable params: 37,453
Non-trainable params: 0
_________________________________________________________________
<BatchDataset element_spec=(TensorSpec(shape=(None, 1, 13), dtype=tf.int32, name=None), TensorSpec(shape=(None, 13), dtype=tf.int32, name=None))>
Epoch 1/100
2022-03-28 05:19:04.116345: I tensorflow/stream_executor/cuda/cuda_dnn.cc:368] Loaded cuDNN version 8100
1/1 [==============================] - 8s 8s/step - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 0.0000e+00 - val_accuracy: 1.0000
Epoch 2/100
1/1 [==============================] - 0s 38ms/step - loss: 0.0000e+00 - accuracy: 1.0000 - val_loss: 0.0000e+00 - val_accuracy: 1.0000
Assume each windows consume about 13 level of the input :
batched_features = tf.constant( [ [ 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0 ], [ 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0 ], ], shape=( 2, 1, 13 ) )
batched_labels = tf.constant( [[ 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0 ], [ 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0 ]], shape=( 2, 13 ) )
Adding more windows is easy by
batched_features = tf.constant( [ [ 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0 ], [ 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0 ], [ 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0 ], ], shape=( 3, 1, 13 ) )
batched_labels = tf.constant( [[ 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0 ], [ 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0 ], [ 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0, 0 ]], shape=( 3, 13 ) )
dataset = tf.data.Dataset.from_tensor_slices((batched_features, batched_labels))
dataset = dataset.batch(10)
batched_features = dataset
At least you tell me what is the purpose they can use reverse windows to have certain results. ( Apmplitues frequency )
The results will look like these for each windows :
[ Output ] : 2 and 3 Windows
# Sequence types with timestep #1:
# <BatchDataset element_spec=(TensorSpec(shape=(None, 1, 13), dtype=tf.int32, name=None), TensorSpec(shape=(None, 13), dtype=tf.int32, name=None))>
# Sequence types with timestep #2:
# <BatchDataset element_spec=(TensorSpec(shape=(None, 1, 13), dtype=tf.int32, name=None), TensorSpec(shape=(None, 13), dtype=tf.int32, name=None))>
[ Result ]:
Which plotting library does this plot belong to?
I'm not very good at ploting grpahs but this is how I came up with what I was looking for:
# Import libraries
import matplotlib.pyplot as plt
import pandas as pd
import seaborn as sns
import numpy as np
data_dict = {'Best fit': [395.0, 401.0, 358.0, 443.0, 357.0, 378.0, 356.0, 356.0, 403.0, 380.0, 397.0, 406.0, 409.0, 414.0, 350.0, 433.0, 345.0, 376.0, 374.0, 379.0, 9.0, 13.0, 10.0, 13.0, 16.0, 12.0, 6.0, 11.0, 20.0, 10.0, 12.0, 11.0, 15.0, 11.0, 11.0, 11.0, 15.0, 10.0, 8.0, 18.0, 864.0, 803.0, 849.0, 858.0, 815.0, 856.0, 927.0, 878.0, 834.0, 837.0, 811.0, 857.0, 848.0, 869.0, 861.0, 820.0, 887.0, 842.0, 834.0, np.nan], 'MDP': [332, 321, 304, 377, 304, 313, 289, 314, 341, 321, 348, 334, 361, 348, 292, 362, 285, 316, 291, 318, 3, 6, 5, 5, 4, 5, 4, 3, 8, 6, 4, 0, 8, 1, 4, 0, 9, 5, 3, 8, 770, 770, 819, 751, 822, 842, 758, 825, 886, 830, 774, 839, 779, 821, 812, 850, 822, 786, 874, 831], 'Q-Learning': [358, 329, 309, 381, 302, 319, 296, 315, 343, 318, 338, 336, 360, 357, 299, 363, 287, 337, 301, 334, 3, 6, 5, 5, 4, 5, 4, 3, 8, 6, 4, 0, 8, 1, 4, 0, 9, 5, 3, 8, 771, 833, 757, 837, 831, 784, 806, 890, 843, 775, 838, 776, 824, 830, 834, 827, 791, 868, 816, 806], 'parametrized_factor': [1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 1.0, 0.2, 0.2, 0.2, 0.2, 0.2, 0.2, 0.2, 0.2, 0.2, 0.2, 0.2, 0.2, 0.2, 0.2, 0.2, 0.2, 0.2, 0.2, 0.2, 0.2, 2.0, 2.0, 2.0, 2.0, 2.0, 2.0, 2.0, 2.0, 2.0, 2.0, 2.0, 2.0, 2.0, 2.0, 2.0, 2.0, 2.0, 2.0, 2.0, 2.0]}
data2 = pd.DataFrame(data_dict)
# figure size
plt.figure(figsize=(12, 8))
ax = sns.pointplot(y="Best fit", x="parametrized_factor", data=data2, linestyles='-.', color='g', capsize=.1, scale=.2, errwidth=.5)
ax = sns.pointplot(y="MDP", x="parametrized_factor", data=data2, linestyles='-', color='r', capsize=.12, scale=.2, errwidth=.5)
ax = sns.pointplot(y="Q-Learning", x="parametrized_factor", data=data2, linestyles=':', color='k', capsize=.15, scale=.5, errwidth=.5)
ax.legend(bbox_to_anchor=(1.05, 1), loc='upper left')
labels = ax.legend(['BestFit', 'MDP', 'Q- Learning']), colors = ['green', 'red', 'black']
i = 0
for l in labels.get_texts():
l.set_color(colors[i])
i+=1
# ADDED: Remove labels.
ax.set_ylabel('Rejection rate')
ax.set_xlabel('Parametrized factor')
plt.setp(ax.get_legend().get_texts(), fontsize='12') # for legend text
plt.setp(ax.get_legend().get_title(), fontsize='12') # for legend title
plt.show()
I still have not figure out how to change the color of the marks on the legend.
I have a tensor 'a', I want to modify a element of it.
a = tf.convert_to_tensor([[1.0, 1.0, 1.0],
[1.0, 2.0, 1.0],
[1.0, 1.0, 1.0]], dtype=tf.float32)
And I can got the index of that element.
index = tf.where(a==2)
How to derive 'b' from 'a'?
b = tf.convert_to_tensor([[1.0, 1.0, 1.0],
[1.0, 0.0, 1.0],
[1.0, 1.0, 1.0]], dtype=tf.float32)
I know that I can't not modify a tensor from this post.
I solve it by using tf.sparse_to_dense()
import tensorflow as tf
a = tf.convert_to_tensor([[1.0, 1.0, 1.0],
[1.0, 2.0, 1.0],
[1.0, 1.0, 1.0]], dtype=tf.float32)
index = tf.where(a > 1)
zero = tf.sparse_to_dense(index, tf.shape(a, out_type=tf.int64), 0., 1.)
update = tf.sparse_to_dense(index, tf.shape(a, out_type=tf.int64), 0., 0.)
b = a * zero + update
with tf.Session() as sess:
print sess.run(b)
Need to plot a confusion matrix with this script. By running it an empty plot appears. Seems I am close to solution. Any hint?
from numpy import *
import matplotlib.pyplot as plt
from pylab import *
conf_arr = [[50.0, 0.0, 0.0, 0.0, 0.0, 0.0, 0.0], [3.0, 26.0, 0.0, 0.0, 0.0, 1.0, 0.0], [0.0, 0.0, 10.0, 0.0, 0.0, 0.0, 0.0], [4.0, 1.0, 0.0, 5.0, 0.0, 0.0, 0.0], [3.0, 0.0, 1.0, 0.0, 6.0, 0.0, 0.0], [0.0, 0.0, 0.0, 0.0, 0.0, 47.0, 0.0], [2.0, 0.0, 0.0, 0.0, 0.0, 0.0, 8.0]]
norm_conf = []
for i in conf_arr:
a = 0
tmp_arr = []
a = sum(i,0)
for j in i:
tmp_arr.append(float(j)/float(a))
norm_conf.append(tmp_arr)
plt.clf()
fig = plt.figure()
ax = fig.add_subplot(111)
res = ax.imshow(array(norm_conf), cmap=cm.jet, interpolation='nearest')
cb = fig.colorbar(res)
savefig("confmat.png", format="png")
Thanks, I have the plot. Now, the ticks in the x-axes are very small (the graph dimension is: 3 cm x 10 cm or so). How can I enlarge them in order to have a more proportioned graph, lets say 10cm x 10 cm plot? A possible reason is that I visualize the graph as a subplot? Was not able to find the suitable literature to adjust that.
You don't need to clear a current figure (plt.clf()) before adding a new one.
#plt.clf() # <<<<< here
fig = plt.figure()
ax = fig.add_subplot(111)