I have many csv files that has multiple rows and columns which are mostly floating point numbers (some are categorical but one-hot encoded).
Each csv file is the representation of one training example.It contains dependent and independent variables in the same file.
(for example, its not like machine learning problem where each row contains all the information and predicts y1, y2,y3 of that row, its like all the rows combined of x1 to x8
will predict all rows combined of y1 to y3. Hence each csv becomes one training example.
representation of one such csv
** The above image is the representation of one of such csv files
Please note that the length/size of each csv varies.
I want to build a simple ann or any other neural net model. I have problem in processing input data. As each csv is one single training example, in which format should i have to store data to pass to a neural net.
Thanks in advance,
skw
Let's say you have some .csv file all with same data format stored in a folder data.
First you can use glob to read the filenames and use pandas to read the csv and convert to numpy array.
import glob
import pandas as pd
csv = [] # read as numpy array
for f in glob.glob('path/*.csv'):
csv.append(pd.read_csv(f).to_numpy)
print(csv[0].shape)
# it should print (num_rows_csv, 11) # as, 11 columns
# now, first 8 columns are features, and last 3 columns are response
X = []
y = []
for arr in csv:
X.append(arr[0:8])
y.append(arr[8:])
X = np.array(X)
y = np.array(y)
Now, it's easy to train this with CNN, LSTM, any model you want.
Related
I have started using PowerBI and am using Python as a data source with the code below. The source data can be downloaded from here (it's about 700 megabytes). The data is originally from here (contained in IOT_2019_pxp.zip).
import pandas as pd
import numpy as np
import os
path = /path/to/file
to_chunk = pd.read_csv(os.path.join(path,'A.txt'), delimiter = '\t', header = [0,1], index_col = [0,1],
iterator=True, chunksize=1000)
def chunker(to_chunk):
to_concat = []
for chunk in to_chunk:
try:
to_concat.append(chunk['BG'].loc['BG'])
except:
pass
return to_concat
A = pd.concat(chunker(to_chunk))
I = np.identity(A.shape[0])
L = pd.DataFrame(np.linalg.inv(I-A), index=A.index, columns=A.columns)
The code simply:
Loads the file A.txt, which is a symmetrical matrix. This matrix has every sector in every region for both rows and columns. In pandas, these form a MultiIndex.
Filters just the region that I need which is BG. Since it's a symmetrical matrix, both row and column are filtered.
The inverse of the matrix is calculated giving us L, which I want to load into PowerBI. This matrix now just has a single regular Index for sector.
This is all well and good however when I load into PowerBI, the first column (sector names for each row i.e. the DataFrame Index) disappears. When the query gets processed, it is as if it were never there. This is true for both dataframes A and L, so it's not an issue of data processing. The column of row names (the DataFrame index) is still there in Python, PowerBI just drops it for some reason.
I need this column so that I can link these tables to other tables in my data model. Any ideas on how to keep it from disappearing at load time?
For what it's worth, calling reset_index() removed the index from the dataframes and they got loaded like regular columns. For whatever reason, PBI does not properly load pandas indices.
For a regular 1D index, I had to do S.reset_index().
For a MultiIndex, I had to do L.reset_index(inplace=True).
I am working with a large tabular dataset that consists of many categorical columns. I want to train a regression model (XGBoost) in this data while using as many regressors as possible.
Because of the size of data, I am using incremental training - where following sklearn API - .fit(X, y) I am not able to fit the entire matrix X into memory and therefore I am training the model in a couple of rows at the time. The problem is that in every batch, the model is expecting the same number of columns in X.
This is where it gets tricky because some variables are categorical it may be that one-hot encoding on a batch of data will same some shape (e.g. 20 columns). However, the next batch will have (26 columns) simply because in the previous batch not every unique level of the categorical feature was present. Sklearn allows for accounting for this and costume function can also be used: To keep some number of columns in matrix X.
import seaborn as sns
import numpy as np
from sklearn.preprocessing import OneHotEncoder
def one_hot_known(dataf, list_levels, col):
"""Creates a dummy coded matrix with as many columns as unique levels"""
return np.array(
[np.eye(len(list_levels))[list_levels.index(i)] for i in dataf[col]])
# Load Some Dataset with categorical variable
df_orig = sns.load_dataset('tips')
# List of unique levels - known apriori
day_level = list(df_orig['day'].unique())
# Image, we have a batch of data (subset of original data) and one categorical level (DAY) is not present here
df = df_orig.loc[lambda d: d['day'] != 'Sun']
# Missing category is filled with 0 and in next batch, if present its columns will have 1.
OneHotEncoder(categories = [day_level], sparse=False).fit_transform(np.array(df['day']).reshape(-1, 1))
#Costum function, can be used in incremental(data batches chunk fashion)
one_hot_known(df, day_level, 'day')
What I would like to do not is to utilize the TargerEncoding approach, so that we do not have matrix X with a huge number of columns. However, it still needs to be done in an Incremental fashion, just like the OneHot Encoding above.
I am writing this as a post because I know this is very useful to many people and would like to know how to utilize the same strategy for TargetEncoding.
I am aware that Deep Learning allows for Embedding layers, which represent categorical features in continuous space but I would like to apply TargetEncoding.
I have a data set of TFRecord files of serialized TensorFlow Example protocol buffers with one Example proto per note, downloaded from https://magenta.tensorflow.org/datasets/nsynth. I am using the test set, which is approximately 1 Gb, in case someone wants to download it, to check the code below. Each Example contains many features: pitch, instrument ...
The code that reads in this data is:
import tensorflow as tf
import numpy as np
sess = tf.InteractiveSession()
# Reading input data
dataset = tf.data.TFRecordDataset('../data/nsynth-test.tfrecord')
# Convert features into tensors
features = {
"pitch": tf.FixedLenFeature([1], dtype=tf.int64),
"audio": tf.FixedLenFeature([64000], dtype=tf.float32),
"instrument_family": tf.FixedLenFeature([1], dtype=tf.int64)}
parse_function = lambda example_proto: tf.parse_single_example(example_proto,features)
dataset = dataset.map(parse_function)
# Consuming TFRecord data.
dataset = dataset.shuffle(buffer_size=10000)
dataset = dataset.batch(batch_size=3)
dataset = dataset.repeat()
iterator = dataset.make_one_shot_iterator()
batch = iterator.get_next()
sess.run(batch)
Now, the pitch ranges from 21 to 108. But I want to consider data of a given pitch only, e.g. pitch = 51. How do I extract this "pitch=51" subset from the whole dataset? Or alternatively, what do I do to make my iterator go through this subset only?
What you have looks pretty good, all you're missing is a filter function.
For example if you only wanted to extract pitch=51, you should add after your map function
dataset = dataset.filter(lambda example: tf.equal(example["pitch"][0], 51))
I have a reporting sheet in excel that contains a set of datapoints that I want to compile from multiple files with the same format into a master dataset.
The initial step I undertook was to extract the data points I need from multiple sheet into one pandas dataframe. See the steps below
I initally imported the excel file and parsed it
import pandas as pd
xl = pd.ExcelFile(r"C:\Users\Nicola\Desktop\ISP 2016-20 Ops-Technical Form.xlsm")
df = xl.parse("FSL, WASH, DRM") #name of sheet #1
Then I located the data points needed for synthesis
a=df.iloc[5:20,3:5]
a1=df.iloc[6:9,10:12]
b=df.iloc[31:35,3:5]
b1=df.iloc[31:35,10:12]
Then I concatenated and equalised columns positioning to maintain the whole list of values within the same column:
dfcon=pd.concat(([a,b]))
dfcon2=pd.concat(([a1,b1]))
new_cols = {x: y for x, y in zip(dfcon.columns, dfcon2.columns)}
dfcont2=dfcon2.append(dfcon.rename(columns=new_cols))
And lastly created a dataframe with the string of values I need
master=pd.DataFrame(dfcont2)
finalmaster=master.transpose()
The next two steps I wish to pursue are:
1) Replicate the same code for 50 excel files
2) Compile all string of values from this set of excel files into one single pandas dataframe without running this code over again and compile manually by exporting it into excel.
Any support would be greatly appreciated. Thanks
I believe need loop by file names created by glob and last concat together (all files have same structure):
import glob
dfs = []
for f in glob.glob('*.xlsm'):
df = pd.read_excel(io=f, sheet_name=1)
a=df.iloc[5:20,3:5]
a1=df.iloc[6:9,10:12]
b=df.iloc[31:35,3:5]
b1=df.iloc[31:35,10:12]
dfcon=pd.concat(([a,b]))
dfcon2=pd.concat(([a1,b1]))
new_cols = {x: y for x, y in zip(dfcon.columns, dfcon2.columns)}
dfcont2=dfcon2.append(dfcon.rename(columns=new_cols))
dfs.append(dfcont2.T)
out = pd.concat(dfs, ignore_index=True)
Found the solution that works for me, thank you for the input, jezrael.
To further explain:
1) Imported the files with same structure from my Desktop directory, parsed and selected the Excel sheet from which data can be extracted from different locations (iloc)
import glob
dfs = []
for f in glob.glob('C:/Users/Nicola/Desktop/OPS Form/*.xlsm'):
df = pd.ExcelFile(io=f, sheet_name=1)
df = df.parse("FSL, WASH, DRM")
a=df.iloc[5:20,3:5]
a1=df.iloc[7:9,10:12]
b=df.iloc[31:35,3:5]
b1=df.iloc[31:35,10:12]
c=df.iloc[50:56,3:5]
c1=df.iloc[38:39,10:12]
d=df.iloc[57:61,3:5]
e=df.iloc[63:71,3:5]
2) Concatenated and repositioned columns order to compose the first version of the dataframe (output)
dfcon=pd.concat(([a,b,c,d,e]))
dfcon2=pd.concat(([a1,b1,c1]))
new_cols = {x: y for x, y in zip(dfcon.columns, dfcon2.columns)}
dfcont2=dfcon2.append(dfcon.rename(columns=new_cols))
dfs.append(dfcont2.T)
3) Output presented the same string of values but repeated twice [same label and form-specific entry] from recursive data pull-outs linked to iloc locations.
output = pd.concat(dfs, ignore_index=True)
4) This last snippet simply allowed me to extract the label only once and to select all entries ordered in odd numbers. With the last concatenation, I generated the dataframe I seeked, ready to be processed analytically.
a=output[2:3]
b=output[1::2]
pd.concat([a,b], axis=0, ignore_index=True)
I want to divide my dataset into train and test sets using stratified sampling(scikitlearn).my approach is as follows :
1) I'am reading a CSV file and loading it using pandas readCSV.so ultimately i'am storing the loaded csv in a dataframe names "dataset"
dataset = pd.readCSV('CSV_NAME)
2) Now i'am applying stratified sampling as :
train,test = train_test_split(dataset,test_size=0.20,stratify=True)
But it throwing the following error :
TypeError: Singleton array array(True, dtype=bool) cannot be considered a valid collection.
So please suggest me the correct way of doing to it.
'train_test_split' needs to know what the target variable is. Therefore, you should change your call to something like:
train,test = train_test_split(dataset[needed columns], dataset.target,test_size=0.20,stratify=True)
Btw, there is a missing single quote in your first line of code.
You could convert the pandas dataframe to a numpy array by the following
import numpy
dataset = pd.readCSV('CSV_NAME')
dataset = array(dataset)
like suggested in the second answer here: https://www.quora.com/How-does-python-pandas-go-along-with-scikit-learn-library-Has-anyone-doing-data-analysis-using-pandas-and-then-then-fit-models-using-scikit-learn
Or you could read the dataset into a numpy array directly.