Compare commits

...

10 Commits

Author SHA1 Message Date
nthu1080112244 b69fc890b6 [debug] 2022-09-02 15:38:44 +08:00
108000207 d022e72ff1 [update] need to try execute_batch for optimize 2022-08-27 16:51:37 +08:00
108000207 1e82af9ee2 [update] use stream type as api 2022-08-25 16:31:01 +08:00
108000207 327e57e72f [update] api for get_csv, fix bugs for large file 2022-08-24 17:42:53 +08:00
108000207 5d5bf0ee51 [update] differential done 2022-08-23 17:51:21 +08:00
108000207 cf83afc8f3 [update] integral functional done 2022-08-23 17:02:23 +08:00
108000207 c8d0248da1 [update] log functional done 2022-08-23 15:32:00 +08:00
108000207 830fde89a3 [update] support multi "***" and export csv 2022-08-23 13:48:17 +08:00
108000207 fcc76c835c [update] sql get DB and build dataFrame 2022-08-23 11:07:01 +08:00
peterlu14 2937043f78 Merge branch 'feature/project_setting' into release/v1.0.0_project 2022-05-06 10:53:29 +08:00
8 changed files with 56967 additions and 0 deletions
File diff suppressed because it is too large Load Diff
File diff suppressed because it is too large Load Diff
File diff suppressed because it is too large Load Diff
File diff suppressed because it is too large Load Diff
+143
View File
@@ -0,0 +1,143 @@
import psycopg2
import pandas as pd
import numpy as np
from os.path import exists
import sys
import os
import time
def log(df, channel):
# TODO: not sure need use 32 or 64? => 64, since large integral will cause overflow
df[f'{channel}_log'] = np.log10(np.absolute(df[channel]))
return df
def integral(df, x_channel, y_channel, pre_last_x, pre_last_y, integral_result):
offset = df[x_channel]
offset = offset.shift(periods=1)
offset[0] = pre_last_x
x_diff = df[x_channel] - offset
# x_diff = df[x_channel].astype(np.int64) - offset.astype(np.int64)
# print('x_diff\n', x_diff)
offset = df[y_channel]
offset = offset.shift(periods=1)
offset[0] = pre_last_y
y_avg = (offset + df[y_channel]) / 2
# y_avg = (offset.astype(np.int64) + df[y_channel].astype(np.int64)) / 2
# print('y_avg:', y_avg)
# print('y_avg: ', y_avg)
local_integral = x_diff * y_avg
# TODO: if user don't need, then can remove it
df[f'{x_channel}_{y_channel}_integral_delta'] = local_integral
local_integral[0] += integral_result
df[f'{x_channel}_{y_channel}_integral'] = local_integral.cumsum()
# print(df)
pre_last_x = df[x_channel].iat[-1]
pre_last_y = df[y_channel].iat[-1]
integral_result = df[f'{x_channel}_{y_channel}_integral'].iat[-1]
# print('next run info: ', pre_last_x, pre_last_y, integral_result)
return df, pre_last_x, pre_last_y, integral_result
def differential(df, x_channel, y_channel, pre_last_x, pre_last_y):
offset = df[x_channel]
offset = offset.shift(periods=1)
offset[0] = pre_last_x
x_diff = df[x_channel] - offset
# x_diff = df[x_channel].astype(np.int64) - offset.astype(np.int64)
offset = df[y_channel]
offset = offset.shift(periods=1)
offset[0] = pre_last_y
y_diff = df[y_channel] - offset
# y_diff = df[y_channel].astype(np.int64) - offset.astype(np.int64)
local_diff = y_diff / x_diff
# TODO: if user don't need, then can remove it
df[f'{x_channel}_{y_channel}_diff'] = local_diff
pre_last_x = df[x_channel].iat[-1]
pre_last_y = df[y_channel].iat[-1]
return df, pre_last_x, pre_last_y
def download(id, mode, x_channel, y_channel):
# TODO: can change different chunck size
pages_chunck_size = 250
sql_total = 0
conn_start = time.time()
# TODO: need the check port id
conn = psycopg2.connect(database="postgres", user="biopro", password="BioProControlBox", host="127.0.0.1", port="54321")
print('conn time: ', time.time() - conn_start)
cursor = conn.cursor()
# recording_data_metas: meta info to get the raw_data table
sql_str = f'SELECT "raw_data" FROM "recording_data_metas" WHERE id = {id}'
first_sql = time.time()
cursor.execute(sql_str)
print('first sql time: ', time.time() - first_sql)
raw_data_list = None
try:
raw_data_list = cursor.fetchall()[0][0]
channels = list(raw_data_list.keys()) # [0,1,2,3]
df_all = pd.DataFrame()
# df = pd.DataFrame()
df_time = []
pre_last_x = 0
pre_last_y = 0
integral_result = 0
print('raw_data_list', raw_data_list)
print('channels', channels)
for page in range(len(raw_data_list[channels[0]])):
df = pd.DataFrame()
print(raw_data_list[channels[0]][page])
for channel in channels:
raw_data_page = raw_data_list[channel][page]
sql_str = f'SELECT "data" FROM "{channel}_recording_data_raws" WHERE id = {raw_data_list[channel][page]}'
sql_start = time.time()
cursor.execute(sql_str)
sql_total += (time.time() - sql_start)
data_value = cursor.fetchall()[0][0].replace('"***"', ' ').split(' ')
if ~('Time' in df.columns):
df['Time'] = data_value[:-1:2]
df[channel] = data_value[1:-1:2]
# df = df.loc[lambda x: x.index % 2 == 1].reset_index(drop=True)
# df.insert(0, 'Time', df_time)
# transfer dataframe from string to number
cols = df.columns
df[cols] = df[cols].apply(pd.to_numeric, errors='coerce') # coerce force non-a-number string be converted to NaN
if mode == 'Log':
df = log(df, x_channel)
elif mode == 'Integral':
df, pre_last_x, pre_last_y, integral_result = integral(df, x_channel, y_channel, pre_last_x, pre_last_y, integral_result)
elif mode == 'Differential':
df, pre_last_x, pre_last_y = differential(df, x_channel, y_channel, pre_last_x, pre_last_y)
df_all = pd.concat([df_all, df], ignore_index=True, sort=False)
print(df_all)
with open(f"../../csv/{id}.csv", mode = 'a') as export_file:
df_all.to_csv(export_file, header=(export_file.tell() == 0))
df_all = pd.DataFrame()
print('total sql: ', sql_total)
except BaseException as e:
print(e)
exit
return
if __name__ == '__main__':
params = {
'id': 0,
'mode': 0,
'x_channel': 0,
'y_channel': 0
}
for input in sys.argv[1:]: # Now we're going to iterate over argv[1:] (argv[0] is the program name)
param = input.split("=") # Get what's left of the '='
params[param[0]] = param[1]
id = params['id']
mode = params['mode']
x_channel = params['x_channel']
y_channel = params['y_channel']
# print('download_functional init at main: ', id, mode, x_channel, y_channel)
start_time = time.time()
# download(id, mode, x_channel, y_channel)
download(849, '', '', '')
# download(408, '', 'Time', '0')
# download(405, 'Integral', 'Time', '0')
# download(405, 'Integral', 'Time', '0')
# download(408, 'Integral', 'Time', '0')
print('total: ', time.time() - start_time)
# download(1262, 'differential', '1', '0')
# 1262/differential/1/0
@@ -1,7 +1,12 @@
import RecordingDataRaw from '../../models/file/recording_data_raw'
import RecordingDataMeta from '../../models/file/recording_data_meta'
import * as auth from '../auth'
import childProcess from 'child_process'
import path from 'path'
import { createReadStream } from 'fs'
const recordingDataRaw = new RecordingDataRaw()
const recordingDataMeta = new RecordingDataMeta()
export const create = async (ctx, next) => {
const index = ctx.params.channel
@@ -162,3 +167,37 @@ export const getAll = async (ctx, next) => {
next()
}
export const getCSV = async (ctx, next) => {
const id = ctx.params.id
const mode = ctx.params.mode
const x_channel = ctx.params.x_channel
const y_channel = ctx.params.y_channel
// console.log('getCSV: ', id, mode, x_channel, y_channel)
// const result = await recordingDataRaw.getCSV(id, mode, x_channel, y_channel)
const process = childProcess.spawnSync('python', [
path.join(__dirname, '/download_functional.py'),
'id=' + id,
'mode=' + mode,
'x_channel=' + x_channel,
'y_channel=' + y_channel,
], { maxBuffer: 2 * 1024 * 1024 * 1024 })
// console.log('process: ', process)
// process.stdout.on('data', (data) => {
// conosle.log('data', data)
// })
const result = await recordingDataMeta.getByID(id)
// console.log('result: ', result)
// console.log('result: ', result[0]['dataValues']['name'])
var file_name = id + '.csv'
// ctx.body = fs.stdout.toString()
// console.log('process.stdout.toString(): ', process.stdout.toString())
// console.log('process.stderr.toString(): ', process.stderr.toString())
ctx.type = 'stream'
ctx.attachment(result[0]['dataValues']['name']+'.csv')
ctx.body = createReadStream(path.resolve(__dirname, '../../csv', file_name))
// need to remove the csv file after download, since using 'append' for set header one time
const fs = require('fs')
// fs.unlinkSync(path.resolve(__dirname, '../../csv', file_name))
next()
}
+60
View File
@@ -0,0 +1,60 @@
time 0 1 2 3 time_0_integral
0 0 -73922 -836265 -848970 0 0.000000e+00
1 10500 -84368 -834380 -848548 0 -8.310225e+08
2 21000 -84389 -834380 -848126 0 -1.716997e+09
3 31500 -84332 -835637 -847704 0 -2.602782e+09
4 42500 -84166 -835637 -847283 0 -3.529521e+09
... ... ... ... ... .. ...
1462 15390500 -45761 -449806 -466674 0 -9.716838e+11
1463 15401000 -45833 -449806 -466445 0 -9.721646e+11
1464 15411500 -45790 -451063 -466216 0 -9.726457e+11
1465 15422500 -45638 -449806 -465988 0 -9.731485e+11
1466 15432500 -45674 -449806 -465759 0 -9.736051e+11
[1467 rows x 6 columns]
next run info: 15432500 -45674 -973605066500.0
time 0 1 2 3 time_0_integral
0 15443000 -45689 -448549 -465531 0 -9.740847e+11
1 15453500 -45689 -448549 -465531 0 -9.745645e+11
2 15464000 -45631 -449178 -465303 0 -9.750439e+11
3 15474500 -45581 -447921 -465075 0 -9.755227e+11
4 15485500 -45646 -447921 -464847 0 -9.760245e+11
... ... ... ... ... .. ...
1443 30720000 -24971 -248122 -259749 0 -1.497382e+12
1444 30730500 -24920 -247950 -259624 0 -1.497644e+12
1445 30741000 -24920 -248381 -259624 0 -1.497905e+12
1446 30751500 -24791 -248036 -259500 0 -1.498166e+12
1447 30762000 -24856 -247346 -259376 0 -1.498427e+12
[1448 rows x 6 columns]
next run info: 30762000 -24856 -1498426872750.0
time 0 1 2 3 time_0_integral
0 30772500 -24913 -247669 -259251 0 -1.498688e+12
1 30783000 -24848 -247669 -259127 0 -1.498949e+12
2 30794000 -24805 -247669 -259003 0 -1.499222e+12
3 30804000 -24805 -247454 -259003 0 -1.499471e+12
4 30814500 -24870 -247238 -258879 0 -1.499731e+12
... ... ... ... ... .. ...
1456 46091500 -13496 -134943 -147002 0 -1.783935e+12
1457 46102000 -13510 -134921 -146934 0 -1.784077e+12
1458 46112500 -13510 -134123 -146934 0 -1.784219e+12
1459 46123000 -13597 -134339 -146866 0 -1.784361e+12
1460 46134000 -13525 -134598 -146799 0 -1.784510e+12
[1461 rows x 6 columns]
next run info: 46134000 -13525 -1784510350500.0
time 0 1 2 3 time_0_integral
0 46144500 -13366 -134210 -146732 0 -1.784652e+12
1 46154500 -13460 -134210 -146665 0 -1.784786e+12
2 46165500 -13503 -134016 -146597 0 -1.784934e+12
3 46176000 -13503 -134102 -146597 0 -1.785076e+12
4 46186500 -13431 -134167 -146530 0 -1.785217e+12
.. ... ... ... ... .. ...
583 52277500 -10487 -105193 -117883 0 -1.858083e+12
584 52288000 -10573 -105042 -117830 0 -1.858194e+12
585 52298500 -10595 -105106 -117778 0 -1.858305e+12
586 52309000 -10494 -105085 -117725 0 -1.858415e+12
587 52319500 -10523 -105085 -117672 0 -1.858526e+12
[588 rows x 6 columns]
next run info: 52319500 -10523 -1858525771000.0
+1
View File
@@ -59,6 +59,7 @@ export default router
.get('/api/file/raw/get_attr_by_id/:channel/:id/:attr', controllers.recordingDataRaw.getAttrByID)
.get('/api/file/raw/get_attr_by_ids/:channel/:id/:attr', controllers.recordingDataRaw.getAttrByIDs)
.get('/api/file/raw/get_attr_by_parent/:channel/:parent/:attr', controllers.recordingDataRaw.getAttrByParent)
.get('/api/file/raw/get_csv/:id/:mode/:x_channel/:y_channel', controllers.recordingDataRaw.getCSV)
// mini
.post('/api/file/mini/create/:channel', controllers.recordingDataMini.create)