blctools 0.0.22

Tools to work with data inside BLCs MS Sharepoint Cloud & Database

BLC Data Analytics tools package

Data Analytics helper functions to work with inside BLC's Cloud system.

Changelog

Version 0.0.22

• TablasVC()

  • Added: One can now initialize the "parques" parameter with a single string, to avoid using brackets when working with just one park.

• DatosCROM()

  • Added: One can now initialize the "parques" parameter with a single string, to avoid using brackets when working with just one park.
  • Fixed bug: Eerror when no active incidents were found for a given parq/period and trying to .interpretar_incidencias_bajo_iec61400()
  • Fixed bug: Some functions where trying to call TablasVC().consultar_incidencias() which doesn't exist anymore. Now its .consultar(incidencias='offline')
  • cargar_incidencias is now False as default.
  • explotar_incidencias is now False as default.

Version 0.0.21

• DatosCROM()
  • Added: .buscar_indisponibilidades(). Searches for periods of time which the WTGs met certain condition filters of out of service
  • Added: .buscar_limitaciones(). Searches for periods of time which the PLANT or WTGs had an SP_P > than the max.
  • Added: .buscar_incidencias() combines both previous functions in a single line.
  • Added: .incidencias_autodetectadas attribute, with export capabilities
  • Added: "Status" column to incidencias_redux
  • Added: "Count" and "Histogram" columns when calculating powercurves. _count "sheet" on export is now always there.
  • Added: "explotar_incidencias" initialization parámeter, which can take either 'iec61400' (default) or 'crudas'
  • Fixed bug: Missing Indices on datos_seg when trying to consolidate incidents and 10second data
  • _identificar_indisponibilidades() pgen and ppos parameters were added, to further improve filtering capacity
  • __procesar_d10s_dc() and cargar_segundales() now use DataFrame.reindex() to fill with blank rows where necessary.
  • elaborar_curvas_de_potencia() now sorts index clockwise instead of alphabetically
• TablasVC()
  • Added .consultar_agrupamiento_de_un_equipo()

Version 0.0.20

• DatosCROM()
  • interpretar_incidencias_bajo_iec61400() made more efficient, while keeping compatibility with pandas >= 1.5.0
  • _identificar_indisponibilidades() pgen and ppos parameters were added, to further improve filtering capacity
• TablasVC()
  • Added "potencia=False" parameter to consultar_agrupamientos(), consultar_equipos_no_agrupamientos()

Version 0.0.19

• DatosCROM()
  • Fixed bug: interpretar_incidencias_bajo_iec61400() was experiencing problems with pandas >= 1.5.0, and had a problem with the index selection statement at the beginning.
  • Added experimentally: _identificar_indisponibilidades(duracion=30)

Version 0.0.18

• DatosCROM()
  • .exportar() now flattens MultiIndex columns, freezes panes and formats floats, for easier view on Excel.
  • Added "invalidar_datos_seg_congelados" initialization parameter (default = False)
  • Fixed bug: .exportar() throwing error when curvas_de_potencia or consolidado held more than 1 Dataframe
  • Fixed bug: .interpretar_incidencias_bajo_iec61400() now trims final result between fecha_i and fecha_f
  • Fixed bug: .elaborar_curvas_de_potencia() wasn't working properly with filtro_n >1

Version 0.0.17

• fechas submodule
  • Added .sem_dia_1(date,offset=0) function. Returns first day of week of the given date + given offset in days (offset can be negative)
  • Added .sem_act_dia_1(offset=0) function. Returns first day of current week + given offset in days
  • Added .año_op(fecha_actual,fecha_cod) function. Returns the current contractual year based on the two dates (First year is 1, not 0)
• DatosCROM()
  • adjusted how xlsx files processing time is calculated
  • adjusted how .pickle files are created/read. Compression was added through bz2
  • consolidado attribute is now called consolidado_redux, which is a subset of columns of consolidado
  • consolidado attribute now holds all the original data, but resampled to the users' input sample rate, excluding categorical variables
  • consultar_equipos_parque() and consultar_equipos_parque_no_agrupamientos() are now called consultar_equipos() and consultar_equipos_no_agrupamientos()
  • __iec61400_preparar_df_vacio_minutal() and it's aux methods are now more efficient
  • Added variables "Gen", "Con", "Pos" as boolean series for each equipment, into the datos_seg dataframe
  • Added columns "ENS_Lim","ENS_LimExt","ENS_LimInt","ENS_Ind","ENS_IndExt,"ENS_IndInt" to "consolidado" report
  • Added variable "Datos_%" which indicates the ammount of not null data per row, per element, into the datos_seg dataframe
  • Added variable "Registros" which indicates the ammout of rows (measured in time) on the excel file for that given element. Should alwasy equal 24
  • Fixed bug: .__procesar_d10s_5i() wasn't ensuring boolean dtypes for 'Gen','Con' and 'Pos', which impacted negatively on .elaborar_reporte_consolidado()
  • Fixed bug: .consolidar_todo() wasn't working with PEGARCIG
  • Fixed bug: .consolidar_todo() wasn't working adding up HOUR columns properly (the rest was fine)
  • Fixed bug: .consolidar_todo() wasn't finishing if there weren't any incidents for a given park on a given time window
  • Fixed bug: .explotar_incidencias() now works properly if there aren't any incidents on a given time window
  • Fixed bug: .exportar() was exporting empty dataframes, since they were not "None"
  • Fixed bug: Roll-up of availability parameters wasn't being done correctly. Now groupings are the mean or logic operator of their groupped elements.
  • Fixed bug: 10s pre-processing: Wind_dir_r was not being saved correctly for 'PLANT' element on self.datos_seg
  • Fixed bug: 10 second data is now being completed (in case of missing 10second records) with null values. This could have lead to serious problems when
  • Fixed bug: On many methods. From pandas >= 1.5.0 numpy series where linked and not copied when creating a new df. This created many bugs throughout the whole program.
  • unloading incidents to the 10 second data, since the main mechanism behind it is selection trhough .loc[t_stamp,:]
• TablasVC()
  • Added attributes .habilitaciones and .tipohabilitaciones which are new tables of "datos básicos". They contain COD related information of each plant
  • Added .consultar_cod(nemo_parque='',uc='') function, to quickly find out about a plant's COD (Birthday) date (grabs from the table above)
  • Added .consultar_año_op(fecha_actual,feca_cod) function. Returns the contractual year (Contractual age) of a power plant
  • Added .__crear_dict_tipo_empresa(). Which fills in thre internal attributes: _empresas_bop, _empresas_gen, _empresas_grid
  • Fixed bug: .consultar() would fetchdatos_basicos from offline file even when 'online' parameter was set. This was due to:
    • a logic error of variable db_off inside .consultar()
    • the order in which instructions were executed
  • Fixed bug: __procesar_incidencias_it2() wasn't executing properly when cargar_datos_basicos = 'offline' was set. Solved by calling the new function .__crear_dict_tipo_empresa() before business mapping
  • Fixed bug: Desconecting twice from DB when fetching incidencias
  • Fixed bug: __combinar_incidencias_locales_y_online() wasn't working properly. Duplicate dropping should only be done by 'ID'
  • Refactored __asignar_datos() .method, to make it modular per table

Version 0.0.16

• DatosCROM()
  • Added data invalidation when processing 10 second data. Now consecutive values of P, Q, Wind, Wind_dir and PPos != 0 and != max (for Ppos) are replaced with np.nan
  • Fixed bug when selecting no .parques ( [] ) and initializing a DatosCROM object
  • Fixed bug when truying to export specific reports through the generating functions (ie elaborar_curvas_de_potencia, interpretar_bajo_iec61400, etc.) and seeting the "exportar" parameter to True
  • Fixed bug. PI wasn't being correctly calculated. (* instead of / was being used)
  • Added "filtro_n" parameter to .elaborar_curvas_de_potencia(), to only show results with an n_count >= n per bin. Default 0 (deactivated)
  • Added feature. Ppos is being calculated if column doesn't exist, based on Curvas_de_potencia.xlsx on BLC's cloud under blc.get_dc_cfg() directory
  • "clientes" initializing parameter was removed (little use for it)
  • "solo_CROM" initializing parameter was removed (little use for it)
• TablasVC()
  • Improved the processing of incidents, now "Origin" is "NA" and np.nan resistent. Default = "INT"
  • Fixed bug through which a filter for discarded incidents wasn't working.
  • Refactored processing of incidents, it's more readable now

Version 0.0.15

• DatosCROM()
  • .exportar() was added. Helper function to export all loaded reports to xlsx format.
  • .explotar_incidencias() now takes in optional export parameters.
  • .incidencias_interpretar_bajo_iec61400() now takes in optional export parameters.
  • .incidencias_identificar_solapamientos() was added.
  • "curvas_de_potencia" parameter added while initializing, to force power curves to be computed automatically
  • .elaborar_curvas_de_potencia() is now more resistent to failures. Powercurves can be elaborated without having to process incidents before, although not recommended if user doesn't know what he is doing.
  • changed how __renombrar_lvl0() works, now it must be feed the equipment names.
• TablasVC()
  • fixed a bug, where .consultar_equipos_parque_no_agrupamientos would return "PLANT" twice for GRIOEO
  • "cargar_incidencias" and "cargar_datos_basicos" parameters added while initializing, with possible values being [Flase, True, 'offline', 'online']
  • now queries per default active incidents from the 1st day of two months prior to today (if True or 'online' values where passed to "cargar_incidencias" parameter). Then concatenates obtained results with offline results (if there are any).

Version 0.0.14

• TablasVC()
  • Fixed bug: Corrected functions that took equipments from blctools.TablasVC().conjuntogeneradores, now reading from .tipoequipo
  • Fixed bug: Now Parks without any grouping of it's generators can return the generators without grouping. Special Case: PEGARCIG
• SQLConnector()
  • now decrypts login data from BLC's secure fileserver.
• DatosCROM()
  • .consultar_equipos_parque() now retrieves the right elements if the plant doesn't havy any other grouping stages
  • .consultar_equipos_por_agrupamiento() now returns an empty list if the plant doesn't havy any other grouping stages
  • .rpt_consolidado(granularidad='1D') was added.
  • "forzar_reprocesamiento" parameter added while initializing, to force re-process 10 second data.
  • .rpt_curvas_de_potencia() was added
  • changed the behaviour of .fecha_i and .fecha_i setters. They now attempt to update "incidencias" and "archivos" depending on the values of fecha_i and fecha_f, but the setters dont try to order the dates automatically, which lead to an unwanted behaviour.
  • Changed completely how DatosCROM() now proceses raw data (10 seconds and Incidencias). It's more efficient, and the heavy lifting occurs while processing the 10-second data. Now index has 1 column (Time) and columns have a 3 level structure (p,e,v) = Park, Element, Variable

Version 0.0.13

• Created SQLConnector class to migrate to SQLAlchemy
• blctools.TablasVC now inherits from SQLConnector

Version 0.0.12

• Function blctools.fechas.obtener_periodo() was added (moved from blctools.ReporteBase())
• Added more parameter flexibility within all functions inside blctools.fechas.
• Objects PPO(), DTE() and DatosCROM() now accept the "periodo" initialization parameter (overrides fecha_i and fecha_f).
• bjects blctools.DatosCROM() and blctools.TablasVC() have now the "solo_CROM=False" initialization parameter.
• function blctools.TablasVC.consultar_nemoCammesa() has now the "solo_CROM=False" argument

Version 0.0.11

• Corrected blctools.PPO.periodo('mes_anterior')
• Function ReporteBase.a_excel() accepts now the parameter "exportar_consulta=bool"
• Objects PPO() and DTE() now accept the following initialization parameters
  • "filtro"
  • "periodo" (overrides fecha_i and fecha_f)
• No more unnecessary warnings on changing the "filtro" parameter
• Function blctools.fechas.mes_dia_1() and blctools.fechas.mes_ult_dia() now validate dates using blctools.fechas.validar_fechas()
• Functions blctools.fechas.validar_fecha() and .validar_fechas() now have a parameter "prevenir_futuro=True" as default, to prevent future dates or not.
• Fixed a bug, when changing "filtro" on functions of child classes of DTE() or PPO(), download and extraction directories wouldn't update properly.

Version 0.0.10

• Attribute blctools.DatosCROM.dir_salida sets to which directory the reports go out to (didn't have much impact before)
• Improved automatic Servicios directory search

Version 0.0.9

• Removed dependencies list
• Ability to download unfiltered PPOs/DTEs without specifying parks/plants list
• Incidents are no longer loaded by default on blctools.DatosCrom() objects

Version 0.0.8

• Corrected dependencies list

Version 0.0.6 ~ 0.0.7

• Bug fixes regarding file management
• Ability to customize filters when hitting CAMMESA's API
• PBA calculation according to IEC61400 is still functioning incorrectly

Version 0.0.5

• Most of the code is Object Oriented now.
• CAMMESA's forecasts have been added.

Version 0.0.2 ~ 0.0.4

Fixed the install issues

Version 0.0.1

• Basic functionality is up and running

TO DOs

• Prevenir que se intenten procesar archivos excel 10 segundales que no tengan datos
• Corroborar que la cantidad de grupos de columnas en los archivos 10 segundales, coincida con la cantidad de equipos del parque cuyo nombre de archivo enuncia.
• Convertir DatosCROM().parques a "nemos" automáticamente.
• Poder tomar mediciones SMEC de la BD del VC
• Mejorar los mensajes de procesamiento nivel 0, 1 y 2 en todas las funciones
• Fix Bug: Code could break by passing an '' element into parques initialization parameter