# Import necessary packages.
import pandas as pd
import numpy as np
import math
import time
from sklearn.ensemble import RandomForestRegressor
import matplotlib.pyplot as plt
import copy- Examine the distribution of the predicted track and validation track.
- Setting up the threshold for calculating the MSE specifically for peaks.
- Creating Models for each histone marks.
# Reading data.
ml_df = pd.read_csv("sources/ML_model/output/ml_data.csv", header=0)
# Create the shuffled dataframe for randomly selecting the folds.
ml_df_shuf = ml_df.sample(frac=1)
# Define number of folds.
k = 4
# Create index for the folds.
folds_index = list(range(0, ml_df_shuf.shape[0], math.ceil(ml_df_shuf.shape[0]/k))) + [ml_df_shuf.shape[0]]
folds_index = [[folds_index[i]+1, folds_index[i+1]] for i in range(len(folds_index)-1)]
folds_index[0][0] = 0
folds_index[[0, 303615], [303616, 607230], [607231, 910845], [910846, 1214460]]
ml_df
<style scoped>
.dataframe tbody tr th:only-of-type {
vertical-align: middle;
}
</style>
.dataframe tbody tr th {
vertical-align: top;
}
.dataframe thead th {
text-align: right;
}
| avo | cur | cell | mark | ideas | valid | |
|---|---|---|---|---|---|---|
| 0 | 0.246995 | 1.801744e-01 | C46 | M03 | 0 | 0.226440 |
| 1 | 0.248117 | 1.801744e-01 | C46 | M03 | 92 | 0.226440 |
| 2 | 0.248174 | 1.801744e-01 | C46 | M03 | 0 | 0.226440 |
| 3 | 0.247927 | 1.801744e-01 | C46 | M03 | 0 | 0.226440 |
| 4 | 0.248312 | 1.801744e-01 | C46 | M03 | 0 | 0.226440 |
| 5 | 0.247233 | 1.801744e-01 | C46 | M03 | 19 | 0.226440 |
| 6 | 0.247316 | 1.801744e-01 | C46 | M03 | 19 | 0.226440 |
| 7 | 0.247123 | 1.801744e-01 | C46 | M03 | 43 | 0.226440 |
| 8 | 0.246343 | 1.801744e-01 | C46 | M03 | 19 | 0.226440 |
| 9 | 0.248208 | 1.801744e-01 | C46 | M03 | 0 | 0.226440 |
| 10 | 0.248134 | 1.801744e-01 | C46 | M03 | 0 | 0.226440 |
| 11 | 0.248359 | 1.801744e-01 | C46 | M03 | 0 | 0.226440 |
| 12 | 0.248478 | 1.801744e-01 | C46 | M03 | 0 | 0.226440 |
| 13 | 0.248123 | 1.801744e-01 | C46 | M03 | 0 | 0.226440 |
| 14 | 0.248436 | 1.801744e-01 | C46 | M03 | 0 | 0.226440 |
| 15 | 0.249592 | 1.801744e-01 | C46 | M03 | 0 | 0.226440 |
| 16 | 0.248901 | 1.801744e-01 | C46 | M03 | 0 | 0.226440 |
| 17 | 0.248855 | 1.801744e-01 | C46 | M03 | 0 | 0.226440 |
| 18 | 0.250269 | 1.801744e-01 | C46 | M03 | 0 | 0.226440 |
| 19 | 0.248659 | 1.801744e-01 | C46 | M03 | 0 | 0.226440 |
| 20 | 0.248206 | 1.801744e-01 | C46 | M03 | 2 | 0.226440 |
| 21 | 0.248728 | 1.801744e-01 | C46 | M03 | 0 | 0.226440 |
| 22 | 0.249281 | 1.801744e-01 | C46 | M03 | 0 | 0.226440 |
| 23 | 0.249216 | 1.801744e-01 | C46 | M03 | 0 | 0.226440 |
| 24 | 0.248470 | 1.801744e-01 | C46 | M03 | 0 | 0.226440 |
| 25 | 0.248940 | 1.801744e-01 | C46 | M03 | 19 | 0.226440 |
| 26 | 0.249946 | 1.801744e-01 | C46 | M03 | 0 | 0.226440 |
| 27 | 0.249419 | 1.801744e-01 | C46 | M03 | 0 | 0.226440 |
| 28 | 0.249438 | 1.801744e-01 | C46 | M03 | 0 | 0.226440 |
| 29 | 0.248860 | 1.801744e-01 | C46 | M03 | 15 | 0.226440 |
| ... | ... | ... | ... | ... | ... | ... |
| 1214430 | 0.182170 | 2.794365e-01 | C23 | M25 | 0 | 0.134385 |
| 1214431 | 0.142648 | 2.398810e-01 | C23 | M25 | 0 | 0.000000 |
| 1214432 | 0.089499 | 7.356942e-02 | C23 | M25 | 0 | 0.000000 |
| 1214433 | 0.085755 | 7.330097e-02 | C23 | M25 | 0 | 0.000000 |
| 1214434 | 0.088690 | 5.963657e-02 | C23 | M25 | 0 | 0.000000 |
| 1214435 | 0.074589 | 5.904327e-02 | C23 | M25 | 0 | 0.000000 |
| 1214436 | 0.071283 | 5.898153e-02 | C23 | M25 | 0 | 0.000000 |
| 1214437 | 0.068884 | 5.900870e-02 | C23 | M25 | 0 | 0.000000 |
| 1214438 | 0.062454 | 6.090621e-02 | C23 | M25 | 0 | 0.000000 |
| 1214439 | 0.066037 | 5.905338e-02 | C23 | M25 | 0 | 0.000000 |
| 1214440 | 0.081037 | 5.905134e-02 | C23 | M25 | 0 | 0.000000 |
| 1214441 | 0.079494 | 5.905811e-02 | C23 | M25 | 0 | 0.000000 |
| 1214442 | 0.079866 | 5.901047e-02 | C23 | M25 | 0 | 0.000000 |
| 1214443 | 0.080911 | 5.904380e-02 | C23 | M25 | 0 | 0.000000 |
| 1214444 | 0.078838 | 5.904843e-02 | C23 | M25 | 0 | 0.000000 |
| 1214445 | 0.064501 | 5.938006e-02 | C23 | M25 | 0 | 0.000000 |
| 1214446 | 0.060386 | 3.004827e-03 | C23 | M25 | 0 | 0.000000 |
| 1214447 | 0.058836 | 6.579292e-04 | C23 | M25 | 0 | 0.000000 |
| 1214448 | 0.057562 | 5.937061e-04 | C23 | M25 | 0 | 0.000000 |
| 1214449 | 0.058535 | 1.503559e-03 | C23 | M25 | 0 | 0.000000 |
| 1214450 | 0.060157 | 3.682951e-07 | C23 | M25 | 0 | 0.000000 |
| 1214451 | 0.062341 | 0.000000e+00 | C23 | M25 | 0 | 0.000000 |
| 1214452 | 0.061212 | 0.000000e+00 | C23 | M25 | 0 | 0.000000 |
| 1214453 | 0.061993 | 0.000000e+00 | C23 | M25 | 0 | 0.000000 |
| 1214454 | 0.061545 | 0.000000e+00 | C23 | M25 | 0 | 0.000000 |
| 1214455 | 0.063010 | 0.000000e+00 | C23 | M25 | 0 | 0.000000 |
| 1214456 | 0.060978 | 0.000000e+00 | C23 | M25 | 0 | 0.000000 |
| 1214457 | 0.061748 | 0.000000e+00 | C23 | M25 | 0 | 0.000000 |
| 1214458 | 0.061063 | 0.000000e+00 | C23 | M25 | 0 | 0.000000 |
| 1214459 | 0.061999 | 0.000000e+00 | C23 | M25 | 0 | 0.000000 |
1214460 rows × 6 columns
# Creating label and feature dataset.
label = ml_df_shuf["valid"].to_numpy()
ml_df_shuf.drop(columns=["valid"])
features_dummies = pd.get_dummies(ml_df_shuf)
features = features_dummies.to_numpy()
print(label[0:5])
print(features[0:5,:])[0.4194066 0.47912769 0.51961 0.499787 0.176605 ]
[[0.468564 0.48123019 0. 0.4194066 0. 0.
0. 0. 0. 0. 0. 1.
0. 0. 0. 1. 0. ]
[0.51875847 0.31617134 0. 0.47912769 0. 0.
0. 0. 0. 0. 1. 0.
0. 0. 1. 0. 0. ]
[0.5656851 0.16588938 0. 0.51961 1. 0.
0. 0. 0. 0. 0. 0.
0. 0. 1. 0. 0. ]
[0.14765898 0.12652433 0. 0.499787 0. 0.
0. 0. 0. 1. 0. 0.
0. 0. 1. 0. 0. ]
[0.18135708 0.17125739 0. 0.176605 0. 0.
0. 0. 0. 0. 0. 0.
0. 1. 0. 0. 1. ]]
def time_stamp():
print("[{}]".format(time.time()))def cv_train(model_name, label, feature, folds):
predicted_set = {}
iter_index = 0
for validation_set in folds:
print("---- Creating Validation Set for data in range {}".format(validation_set))
if model_name == "RF":
regr = RandomForestRegressor()
else:
# Part for adding more models later.
print("#### Invalid Model Name. Terminating Training")
break
# Creating train and valid features, labels.
curr_test_feature = feature[validation_set[0]:validation_set[1],:]
curr_test_label = label[validation_set[0]:validation_set[1]]
curr_train_feature = np.delete(feature, np.s_[validation_set[0]:validation_set[1]+1], axis=0)
curr_train_label = np.delete(label, np.s_[validation_set[0]:validation_set[1]+1], axis=0)
start_time = time.time()
print("-------- Start Training on {}".format(start_time))
regr.fit(curr_train_feature, curr_train_label)
print("-------- Finished Training, elapsed time: {}".format(time.time() - start_time))
predicted_set[iter_index] = {
"model": regr,
"predicted": regr.predict(curr_test_feature),
"test_label": curr_test_label,
"avocado": curr_test_feature[:,0],
"curr_impute": curr_test_feature[:,1]
}
iter_index += 1
return predicted_setdef pretty(d, indent=0):
for key, value in d.items():
print('\t' * indent + str(key))
if isinstance(value, dict):
pretty(value, indent+1)
else:
print('\t' * (indent+1) + str(value))model_dic = cv_train("RF", label, features, folds_index)---- Creating Validation Set for data in range [0, 303615]
-------- Start Training on 1574182987.977485
/Users/Michavillson/anaconda3/lib/python3.7/site-packages/sklearn/ensemble/forest.py:245: FutureWarning: The default value of n_estimators will change from 10 in version 0.20 to 100 in 0.22.
"10 in version 0.20 to 100 in 0.22.", FutureWarning)
-------- Finished Training, elapsed time: 58.52015209197998
---- Creating Validation Set for data in range [303616, 607230]
-------- Start Training on 1574183048.340859
/Users/Michavillson/anaconda3/lib/python3.7/site-packages/sklearn/ensemble/forest.py:245: FutureWarning: The default value of n_estimators will change from 10 in version 0.20 to 100 in 0.22.
"10 in version 0.20 to 100 in 0.22.", FutureWarning)
-------- Finished Training, elapsed time: 58.19038510322571
---- Creating Validation Set for data in range [607231, 910845]
-------- Start Training on 1574183108.437985
/Users/Michavillson/anaconda3/lib/python3.7/site-packages/sklearn/ensemble/forest.py:245: FutureWarning: The default value of n_estimators will change from 10 in version 0.20 to 100 in 0.22.
"10 in version 0.20 to 100 in 0.22.", FutureWarning)
-------- Finished Training, elapsed time: 57.9245810508728
---- Creating Validation Set for data in range [910846, 1214460]
-------- Start Training on 1574183168.262582
/Users/Michavillson/anaconda3/lib/python3.7/site-packages/sklearn/ensemble/forest.py:245: FutureWarning: The default value of n_estimators will change from 10 in version 0.20 to 100 in 0.22.
"10 in version 0.20 to 100 in 0.22.", FutureWarning)
-------- Finished Training, elapsed time: 57.9299750328064
# Calculating the MSE for each fold.
MSE_dict = {}
for i in range(4):
MSE_dict[i] = {
"predicted": round(np.mean(abs(model_dic[i]["predicted"]-model_dic[i]["test_label"])), 4),
"avocado": round(np.mean(abs(model_dic[i]["avocado"]-model_dic[i]["test_label"])), 4),
"curr_impute": round(np.mean(abs(model_dic[i]["curr_impute"]-model_dic[i]["test_label"])), 4)
}
pretty(MSE_dict)0
predicted
0.1372
avocado
0.1537
curr_impute
0.2792
1
predicted
0.1402
avocado
0.1577
curr_impute
0.2832
2
predicted
0.1361
avocado
0.1529
curr_impute
0.277
3
predicted
0.1367
avocado
0.1546
curr_impute
0.2823
# Pairwise Data Visualization.
for i in range(4):
temp_dic = copy.deepcopy(model_dic[i])
temp_dic.pop("model")
temp_plot_df = pd.DataFrame(temp_dic)
temp_plot_df.plot(subplots=True, figsize=(10, 6))
# Plot the feature importance.
feat_imp_li = model_dic[0]["model"].feature_importances_
feat_importance = pd.Series(feat_imp_li, index=features_orig.columns)
feat_importance.nlargest(6).plot(kind='barh')
# print(model_dic[0]["model"].feature_importances_)<matplotlib.axes._subplots.AxesSubplot at 0x1abfdb58d0>
print(features[0,:])[0.2037363 0.25777569 4. 0. 0. 1.
0. 0. 0. 0. 0. 0.
0. 0. 0. 1. ]
ml_df_shuf.head(5)
<style scoped>
.dataframe tbody tr th:only-of-type {
vertical-align: middle;
}
</style>
.dataframe tbody tr th {
vertical-align: top;
}
.dataframe thead th {
text-align: right;
}
| avo | cur | cell | mark | ideas | valid | |
|---|---|---|---|---|---|---|
| 792090 | 0.468564 | 0.481230 | C32 | M21 | 0 | 0.419407 |
| 1151646 | 0.518758 | 0.316171 | C31 | M03 | 0 | 0.479128 |
| 384424 | 0.565685 | 0.165889 | C17 | M03 | 0 | 0.519610 |
| 579290 | 0.147659 | 0.126524 | C27 | M03 | 0 | 0.499787 |
| 956422 | 0.181357 | 0.171257 | C46 | M25 | 0 | 0.176605 |
# Remove all IDEAS State dummies variable.
features_wt_idea = ml_df_shuf[["avo", "cur", "cell", "mark"]]
features_wt_idea = pd.get_dummies(features_wt_idea).to_numpy()
features_wt_idea[0,:]array([0.468564 , 0.48123019, 0. , 0. , 0. ,
0. , 0. , 0. , 0. , 1. ,
0. , 0. , 0. , 1. , 0. ])
model_dic_id_rmd = cv_train("RF", label, features_wt_idea, folds_index)---- Creating Validation Set for data in range [0, 303615]
-------- Start Training on 1574183999.332468
/Users/Michavillson/anaconda3/lib/python3.7/site-packages/sklearn/ensemble/forest.py:245: FutureWarning: The default value of n_estimators will change from 10 in version 0.20 to 100 in 0.22.
"10 in version 0.20 to 100 in 0.22.", FutureWarning)
-------- Finished Training, elapsed time: 56.25135779380798
---- Creating Validation Set for data in range [303616, 607230]
-------- Start Training on 1574184057.5759418
/Users/Michavillson/anaconda3/lib/python3.7/site-packages/sklearn/ensemble/forest.py:245: FutureWarning: The default value of n_estimators will change from 10 in version 0.20 to 100 in 0.22.
"10 in version 0.20 to 100 in 0.22.", FutureWarning)
-------- Finished Training, elapsed time: 55.562814235687256
---- Creating Validation Set for data in range [607231, 910845]
-------- Start Training on 1574184115.145239
/Users/Michavillson/anaconda3/lib/python3.7/site-packages/sklearn/ensemble/forest.py:245: FutureWarning: The default value of n_estimators will change from 10 in version 0.20 to 100 in 0.22.
"10 in version 0.20 to 100 in 0.22.", FutureWarning)
-------- Finished Training, elapsed time: 63.64612078666687
---- Creating Validation Set for data in range [910846, 1214460]
-------- Start Training on 1574184181.19211
/Users/Michavillson/anaconda3/lib/python3.7/site-packages/sklearn/ensemble/forest.py:245: FutureWarning: The default value of n_estimators will change from 10 in version 0.20 to 100 in 0.22.
"10 in version 0.20 to 100 in 0.22.", FutureWarning)
-------- Finished Training, elapsed time: 67.51104378700256
# Calculating the MSE for each fold.
MSE_dict = {}
for i in range(4):
MSE_dict[i] = {
"predicted": round(np.mean(abs(model_dic_id_rmd[i]["predicted"]-model_dic_id_rmd[i]["test_label"])), 4),
"avocado": round(np.mean(abs(model_dic_id_rmd[i]["avocado"]-model_dic_id_rmd[i]["test_label"])), 4),
"curr_impute": round(np.mean(abs(model_dic_id_rmd[i]["curr_impute"]-model_dic_id_rmd[i]["test_label"])), 4)
}
pretty(MSE_dict)0
predicted
0.1371
avocado
0.155
curr_impute
0.2808
1
predicted
0.1385
avocado
0.1552
curr_impute
0.2796
2
predicted
0.1389
avocado
0.1541
curr_impute
0.2796
3
predicted
0.1381
avocado
0.1545
curr_impute
0.2817
ml_df_shuf.head(5)
<style scoped>
.dataframe tbody tr th:only-of-type {
vertical-align: middle;
}
</style>
.dataframe tbody tr th {
vertical-align: top;
}
.dataframe thead th {
text-align: right;
}
| avo | cur | cell | mark | ideas | valid | |
|---|---|---|---|---|---|---|
| 792090 | 0.468564 | 0.481230 | C32 | M21 | 0 | 0.419407 |
| 1151646 | 0.518758 | 0.316171 | C31 | M03 | 0 | 0.479128 |
| 384424 | 0.565685 | 0.165889 | C17 | M03 | 0 | 0.519610 |
| 579290 | 0.147659 | 0.126524 | C27 | M03 | 0 | 0.499787 |
| 956422 | 0.181357 | 0.171257 | C46 | M25 | 0 | 0.176605 |
# Print unique marks.
marks_list = ml_df_shuf.mark.unique()
print(marks_list)['M21' 'M03' 'M25']
def hist_models(mark_id, orig_df, ideas=True):
df = orig_df[orig_df.mark == mark_id]
label = df["valid"].to_numpy()
if ideas:
features = pd.get_dummies(df[["avo", "cur", "cell", "ideas"]]).to_numpy()
else:
features = pd.get_dummies(df[["avo", "cur", "cell"]]).to_numpy()
# Define number of folds.
k = 4
# Create index for the folds.
folds_index = list(range(0, df.shape[0], math.ceil(df.shape[0]/k))) + [df.shape[0]]
folds_index = [[folds_index[i]+1, folds_index[i+1]] for i in range(len(folds_index)-1)]
folds_index[0][0] = 0
folds_index
model_dic = cv_train("RF", label, features, folds_index)
MSE_dict = {}
for i in range(4):
MSE_dict[i] = {
"predicted": round(np.mean(abs(model_dic[i]["predicted"]-model_dic[i]["test_label"])), 4),
"avocado": round(np.mean(abs(model_dic[i]["avocado"]-model_dic[i]["test_label"])), 4),
"curr_impute": round(np.mean(abs(model_dic[i]["curr_impute"]-model_dic[i]["test_label"])), 4)
}
pretty(MSE_dict)
return {
"model_dic": model_dic,
"MSE": MSE_dict
}m21_result = hist_models("M21", ml_df_shuf)---- Creating Validation Set for data in range [0, 116775]
-------- Start Training on 1574188344.497136
/Users/Michavillson/anaconda3/lib/python3.7/site-packages/sklearn/ensemble/forest.py:245: FutureWarning: The default value of n_estimators will change from 10 in version 0.20 to 100 in 0.22.
"10 in version 0.20 to 100 in 0.22.", FutureWarning)
-------- Finished Training, elapsed time: 17.838374853134155
---- Creating Validation Set for data in range [116776, 233550]
-------- Start Training on 1574188362.94751
/Users/Michavillson/anaconda3/lib/python3.7/site-packages/sklearn/ensemble/forest.py:245: FutureWarning: The default value of n_estimators will change from 10 in version 0.20 to 100 in 0.22.
"10 in version 0.20 to 100 in 0.22.", FutureWarning)
-------- Finished Training, elapsed time: 17.883373975753784
---- Creating Validation Set for data in range [233551, 350325]
-------- Start Training on 1574188381.4168372
/Users/Michavillson/anaconda3/lib/python3.7/site-packages/sklearn/ensemble/forest.py:245: FutureWarning: The default value of n_estimators will change from 10 in version 0.20 to 100 in 0.22.
"10 in version 0.20 to 100 in 0.22.", FutureWarning)
-------- Finished Training, elapsed time: 17.50961685180664
---- Creating Validation Set for data in range [350326, 467100]
-------- Start Training on 1574188399.50402
/Users/Michavillson/anaconda3/lib/python3.7/site-packages/sklearn/ensemble/forest.py:245: FutureWarning: The default value of n_estimators will change from 10 in version 0.20 to 100 in 0.22.
"10 in version 0.20 to 100 in 0.22.", FutureWarning)
-------- Finished Training, elapsed time: 17.69806694984436
0
predicted
0.1315
avocado
0.1523
curr_impute
0.3132
1
predicted
0.1315
avocado
0.1493
curr_impute
0.3041
2
predicted
0.1337
avocado
0.1541
curr_impute
0.3129
3
predicted
0.1332
avocado
0.1519
curr_impute
0.3134
m21_result_wt_ideas = hist_models("M21", ml_df_shuf, False)---- Creating Validation Set for data in range [0, 116775]
-------- Start Training on 1574188997.884465
/Users/Michavillson/anaconda3/lib/python3.7/site-packages/sklearn/ensemble/forest.py:245: FutureWarning: The default value of n_estimators will change from 10 in version 0.20 to 100 in 0.22.
"10 in version 0.20 to 100 in 0.22.", FutureWarning)
-------- Finished Training, elapsed time: 16.571380853652954
---- Creating Validation Set for data in range [116776, 233550]
-------- Start Training on 1574189015.0444632
/Users/Michavillson/anaconda3/lib/python3.7/site-packages/sklearn/ensemble/forest.py:245: FutureWarning: The default value of n_estimators will change from 10 in version 0.20 to 100 in 0.22.
"10 in version 0.20 to 100 in 0.22.", FutureWarning)
-------- Finished Training, elapsed time: 16.519545793533325
---- Creating Validation Set for data in range [233551, 350325]
-------- Start Training on 1574189032.130722
/Users/Michavillson/anaconda3/lib/python3.7/site-packages/sklearn/ensemble/forest.py:245: FutureWarning: The default value of n_estimators will change from 10 in version 0.20 to 100 in 0.22.
"10 in version 0.20 to 100 in 0.22.", FutureWarning)
-------- Finished Training, elapsed time: 16.46050000190735
---- Creating Validation Set for data in range [350326, 467100]
-------- Start Training on 1574189049.203674
/Users/Michavillson/anaconda3/lib/python3.7/site-packages/sklearn/ensemble/forest.py:245: FutureWarning: The default value of n_estimators will change from 10 in version 0.20 to 100 in 0.22.
"10 in version 0.20 to 100 in 0.22.", FutureWarning)
-------- Finished Training, elapsed time: 17.15988779067993
0
predicted
0.1319
avocado
0.1523
curr_impute
0.3132
1
predicted
0.1318
avocado
0.1493
curr_impute
0.3041
2
predicted
0.1339
avocado
0.1541
curr_impute
0.3129
3
predicted
0.1332
avocado
0.1519
curr_impute
0.3134
m03_result = hist_models("M03", ml_df_shuf)---- Creating Validation Set for data in range [0, 93420]
-------- Start Training on 1574188493.600275
/Users/Michavillson/anaconda3/lib/python3.7/site-packages/sklearn/ensemble/forest.py:245: FutureWarning: The default value of n_estimators will change from 10 in version 0.20 to 100 in 0.22.
"10 in version 0.20 to 100 in 0.22.", FutureWarning)
-------- Finished Training, elapsed time: 13.731431722640991
---- Creating Validation Set for data in range [93421, 186840]
-------- Start Training on 1574188507.847035
/Users/Michavillson/anaconda3/lib/python3.7/site-packages/sklearn/ensemble/forest.py:245: FutureWarning: The default value of n_estimators will change from 10 in version 0.20 to 100 in 0.22.
"10 in version 0.20 to 100 in 0.22.", FutureWarning)
-------- Finished Training, elapsed time: 12.977031230926514
---- Creating Validation Set for data in range [186841, 280260]
-------- Start Training on 1574188521.262924
/Users/Michavillson/anaconda3/lib/python3.7/site-packages/sklearn/ensemble/forest.py:245: FutureWarning: The default value of n_estimators will change from 10 in version 0.20 to 100 in 0.22.
"10 in version 0.20 to 100 in 0.22.", FutureWarning)
-------- Finished Training, elapsed time: 12.248181104660034
---- Creating Validation Set for data in range [280261, 373680]
-------- Start Training on 1574188533.940921
/Users/Michavillson/anaconda3/lib/python3.7/site-packages/sklearn/ensemble/forest.py:245: FutureWarning: The default value of n_estimators will change from 10 in version 0.20 to 100 in 0.22.
"10 in version 0.20 to 100 in 0.22.", FutureWarning)
-------- Finished Training, elapsed time: 12.40692687034607
0
predicted
0.1064
avocado
0.1181
curr_impute
0.2546
1
predicted
0.1058
avocado
0.1173
curr_impute
0.255
2
predicted
0.1081
avocado
0.1195
curr_impute
0.2601
3
predicted
0.1075
avocado
0.1199
curr_impute
0.2613
m03_result_wt_ideas = hist_models("M03", ml_df_shuf, False)---- Creating Validation Set for data in range [0, 93420]
-------- Start Training on 1574188889.4063501
/Users/Michavillson/anaconda3/lib/python3.7/site-packages/sklearn/ensemble/forest.py:245: FutureWarning: The default value of n_estimators will change from 10 in version 0.20 to 100 in 0.22.
"10 in version 0.20 to 100 in 0.22.", FutureWarning)
-------- Finished Training, elapsed time: 12.069170951843262
---- Creating Validation Set for data in range [93421, 186840]
-------- Start Training on 1574188901.90952
/Users/Michavillson/anaconda3/lib/python3.7/site-packages/sklearn/ensemble/forest.py:245: FutureWarning: The default value of n_estimators will change from 10 in version 0.20 to 100 in 0.22.
"10 in version 0.20 to 100 in 0.22.", FutureWarning)
-------- Finished Training, elapsed time: 11.77048110961914
---- Creating Validation Set for data in range [186841, 280260]
-------- Start Training on 1574188914.099287
/Users/Michavillson/anaconda3/lib/python3.7/site-packages/sklearn/ensemble/forest.py:245: FutureWarning: The default value of n_estimators will change from 10 in version 0.20 to 100 in 0.22.
"10 in version 0.20 to 100 in 0.22.", FutureWarning)
-------- Finished Training, elapsed time: 11.769668817520142
---- Creating Validation Set for data in range [280261, 373680]
-------- Start Training on 1574188926.2890701
/Users/Michavillson/anaconda3/lib/python3.7/site-packages/sklearn/ensemble/forest.py:245: FutureWarning: The default value of n_estimators will change from 10 in version 0.20 to 100 in 0.22.
"10 in version 0.20 to 100 in 0.22.", FutureWarning)
-------- Finished Training, elapsed time: 11.894800901412964
0
predicted
0.1067
avocado
0.1181
curr_impute
0.2546
1
predicted
0.1053
avocado
0.1173
curr_impute
0.255
2
predicted
0.1079
avocado
0.1195
curr_impute
0.2601
3
predicted
0.1077
avocado
0.1199
curr_impute
0.2613
m25_result = hist_models("M25", ml_df_shuf)---- Creating Validation Set for data in range [0, 93420]
-------- Start Training on 1574188633.599432
/Users/Michavillson/anaconda3/lib/python3.7/site-packages/sklearn/ensemble/forest.py:245: FutureWarning: The default value of n_estimators will change from 10 in version 0.20 to 100 in 0.22.
"10 in version 0.20 to 100 in 0.22.", FutureWarning)
-------- Finished Training, elapsed time: 13.147719860076904
---- Creating Validation Set for data in range [93421, 186840]
-------- Start Training on 1574188647.193089
/Users/Michavillson/anaconda3/lib/python3.7/site-packages/sklearn/ensemble/forest.py:245: FutureWarning: The default value of n_estimators will change from 10 in version 0.20 to 100 in 0.22.
"10 in version 0.20 to 100 in 0.22.", FutureWarning)
-------- Finished Training, elapsed time: 12.824733972549438
---- Creating Validation Set for data in range [186841, 280260]
-------- Start Training on 1574188660.440157
/Users/Michavillson/anaconda3/lib/python3.7/site-packages/sklearn/ensemble/forest.py:245: FutureWarning: The default value of n_estimators will change from 10 in version 0.20 to 100 in 0.22.
"10 in version 0.20 to 100 in 0.22.", FutureWarning)
-------- Finished Training, elapsed time: 12.775397062301636
---- Creating Validation Set for data in range [280261, 373680]
-------- Start Training on 1574188673.624012
/Users/Michavillson/anaconda3/lib/python3.7/site-packages/sklearn/ensemble/forest.py:245: FutureWarning: The default value of n_estimators will change from 10 in version 0.20 to 100 in 0.22.
"10 in version 0.20 to 100 in 0.22.", FutureWarning)
-------- Finished Training, elapsed time: 12.997114896774292
0
predicted
0.1729
avocado
0.1953
curr_impute
0.2664
1
predicted
0.1769
avocado
0.2005
curr_impute
0.2736
2
predicted
0.1698
avocado
0.1891
curr_impute
0.2577
3
predicted
0.1713
avocado
0.192
curr_impute
0.2624
m25_result_wt_ideas = hist_models("M25", ml_df_shuf, False)---- Creating Validation Set for data in range [0, 93420]
-------- Start Training on 1574189102.0591948
/Users/Michavillson/anaconda3/lib/python3.7/site-packages/sklearn/ensemble/forest.py:245: FutureWarning: The default value of n_estimators will change from 10 in version 0.20 to 100 in 0.22.
"10 in version 0.20 to 100 in 0.22.", FutureWarning)
-------- Finished Training, elapsed time: 11.923866271972656
---- Creating Validation Set for data in range [93421, 186840]
-------- Start Training on 1574189114.4053
/Users/Michavillson/anaconda3/lib/python3.7/site-packages/sklearn/ensemble/forest.py:245: FutureWarning: The default value of n_estimators will change from 10 in version 0.20 to 100 in 0.22.
"10 in version 0.20 to 100 in 0.22.", FutureWarning)
-------- Finished Training, elapsed time: 11.734225988388062
---- Creating Validation Set for data in range [186841, 280260]
-------- Start Training on 1574189126.537584
/Users/Michavillson/anaconda3/lib/python3.7/site-packages/sklearn/ensemble/forest.py:245: FutureWarning: The default value of n_estimators will change from 10 in version 0.20 to 100 in 0.22.
"10 in version 0.20 to 100 in 0.22.", FutureWarning)
-------- Finished Training, elapsed time: 11.834797859191895
---- Creating Validation Set for data in range [280261, 373680]
-------- Start Training on 1574189138.7640052
/Users/Michavillson/anaconda3/lib/python3.7/site-packages/sklearn/ensemble/forest.py:245: FutureWarning: The default value of n_estimators will change from 10 in version 0.20 to 100 in 0.22.
"10 in version 0.20 to 100 in 0.22.", FutureWarning)
-------- Finished Training, elapsed time: 13.448953866958618
0
predicted
0.1733
avocado
0.1953
curr_impute
0.2664
1
predicted
0.1786
avocado
0.2005
curr_impute
0.2736
2
predicted
0.1724
avocado
0.1891
curr_impute
0.2577
3
predicted
0.1714
avocado
0.192
curr_impute
0.2624
print(marks_list)['M21' 'M03' 'M25']
hist_list = [["Mark_id", "Iteration", "predicted", "avocado", "curr_impute", "ideas"]]
M25_list = [["M25", i[0], i[1]["predicted"], i[1]["avocado"], i[1]["curr_impute"], False] for i in m25_result_wt_ideas["MSE"].items()]
M25_list += [["M25", i[0], i[1]["predicted"], i[1]["avocado"], i[1]["curr_impute"], True] for i in m25_result["MSE"].items()]
M03_list = [["M03", i[0], i[1]["predicted"], i[1]["avocado"], i[1]["curr_impute"], False] for i in m03_result_wt_ideas["MSE"].items()]
M03_list += [["M03", i[0], i[1]["predicted"], i[1]["avocado"], i[1]["curr_impute"], True] for i in m03_result["MSE"].items()]
M21_list = [["M21", i[0], i[1]["predicted"], i[1]["avocado"], i[1]["curr_impute"], False] for i in m21_result_wt_ideas["MSE"].items()]
M21_list += [["M21", i[0], i[1]["predicted"], i[1]["avocado"], i[1]["curr_impute"], True] for i in m21_result["MSE"].items()]
hist_df = pd.DataFrame(M25_list + M03_list + M21_list, columns = hist_list[0])hist_df.to_csv("sources/ML_model/output/mse_hist_marks_model.csv", index=False)
hist_df
<style scoped>
.dataframe tbody tr th:only-of-type {
vertical-align: middle;
}
</style>
.dataframe tbody tr th {
vertical-align: top;
}
.dataframe thead th {
text-align: right;
}
| Mark_id | Iteration | predicted | avocado | curr_impute | ideas | |
|---|---|---|---|---|---|---|
| 0 | M25 | 0 | 0.1733 | 0.1953 | 0.2664 | False |
| 1 | M25 | 1 | 0.1786 | 0.2005 | 0.2736 | False |
| 2 | M25 | 2 | 0.1724 | 0.1891 | 0.2577 | False |
| 3 | M25 | 3 | 0.1714 | 0.1920 | 0.2624 | False |
| 4 | M25 | 0 | 0.1729 | 0.1953 | 0.2664 | True |
| 5 | M25 | 1 | 0.1769 | 0.2005 | 0.2736 | True |
| 6 | M25 | 2 | 0.1698 | 0.1891 | 0.2577 | True |
| 7 | M25 | 3 | 0.1713 | 0.1920 | 0.2624 | True |
| 8 | M03 | 0 | 0.1067 | 0.1181 | 0.2546 | False |
| 9 | M03 | 1 | 0.1053 | 0.1173 | 0.2550 | False |
| 10 | M03 | 2 | 0.1079 | 0.1195 | 0.2601 | False |
| 11 | M03 | 3 | 0.1077 | 0.1199 | 0.2613 | False |
| 12 | M03 | 0 | 0.1064 | 0.1181 | 0.2546 | True |
| 13 | M03 | 1 | 0.1058 | 0.1173 | 0.2550 | True |
| 14 | M03 | 2 | 0.1081 | 0.1195 | 0.2601 | True |
| 15 | M03 | 3 | 0.1075 | 0.1199 | 0.2613 | True |
| 16 | M21 | 0 | 0.1319 | 0.1523 | 0.3132 | False |
| 17 | M21 | 1 | 0.1318 | 0.1493 | 0.3041 | False |
| 18 | M21 | 2 | 0.1339 | 0.1541 | 0.3129 | False |
| 19 | M21 | 3 | 0.1332 | 0.1519 | 0.3134 | False |
| 20 | M21 | 0 | 0.1315 | 0.1523 | 0.3132 | True |
| 21 | M21 | 1 | 0.1315 | 0.1493 | 0.3041 | True |
| 22 | M21 | 2 | 0.1337 | 0.1541 | 0.3129 | True |
| 23 | M21 | 3 | 0.1332 | 0.1519 | 0.3134 | True |