+#I just started to code for some cryptocurrencies analysis. But I just encounterd an issue with the code.
_____________________________________________________________________________
library(magrittr)
library(dplyr)
library(tidyverse)
library(crypto)
library(lubridate)
library(xts)
library(quantmod)
library(tidyr)
df <- getCoins(limit = 50)
yourfunction <- function(df, frequency = NULL) {
freq <- frequency
df$date <- lubridate::round_date(df$date, freq)
data <-
df %>% dplyr::group_by(date, slug, symbol, name, ranknow) %>%
dplyr::summarise(
open = dplyr::first(open),
high = max(high),
low = min(low),
close = dplyr::last(close),
volume = sum(volume),
market = dplyr::last(market))
data$volume <- round(data$volume, digits = 0)
data$market <- round(data$market, digits = 0)
data <- as.data.frame(data)
results <- xts::xts(data[, 2:ncol(data)], as.POSIXct(data[, 1], format =
"%d.%m.%Y %H:%M:%S"))
return(results)
}
#### IF NO ERRORS IN ABOVE RUN THESE INDIVIDUALLY------
week_xts <- yourfunction(df, frequency = "week")
str(week_xts)
as.tbl(week_xts) %>%
select(date, slug, open) %>%
spread(slug, open) %>%
tail()
#### IF NO ERRORS IN ABOVE RUN THESE INDIVIDUALLY------
month_xts <- yourfunction(df, frequency = "month")
str(month_xts)
as.tbl(month_xts) %>%
select(date, slug, open) %>%
spread(slug, open) %>%
tail()
########################
#Making daily, weekly and monthly open for ten biggest coins
daily_open <- spread(df[,c(1,4,6)],slug,open)
weekly_open <- spread(week_xts[,c(1, 2, 6)],slug,open)
monthly_open <- spread(month_xts[,c(1, 2, 6)],slug,open)
ten_biggest <- c(list_of_coins[1:10,"slug"])
daily_open <- daily_open[,c("date",ten_biggest)]
weekly_open <- weekly_open[,c("date",ten_biggest)]
monthly_open <- monthly_open[,c("date",ten_biggest)]
#Making daily, weekly and monthly high for ten biggest coins
daily_high <- spread(df[,c(1,4,7)],slug,high)
weekly_high <- spread(week_xts[,c(1,2,7)],slug,high)
monthly_high <- spread(month_xts[,c(1,2,7)],slug,high)
daily_high <- daily_high[,c("date",ten_biggest)]
weekly_high <- weekly_high[,c("date",ten_biggest)]
monthly_high <- monthly_high[,c("date",ten_biggest)]
#Making daily, weekly and monthly low for ten biggest coins
daily_low <- spread(df[,c(1,4,8)],slug,low)
weekly_low <- spread(week_xts[,c(1,2,8)],slug,low)
monthly_low <- spread(month_xts[,c(1,2,8)],slug,low)
daily_low <- daily_low[,c("date",ten_biggest)]
weekly_low <- weekly_low[,c("date",ten_biggest)]
monthly_low <- monthly_low[,c("date",ten_biggest)]
#Making daily, weekly and monthly close for ten biggest coins
daily_close <- spread(df[,c(1,4,9)],slug,close)
weekly_close <- spread(week_xts[,c(1,2,9)],slug,close)
monthly_close <- spread(month_xts[,c(1,2,9)],slug,close)
daily_close <- daily_close[,c("date",ten_biggest)]
weekly_close <- weekly_close[,c("date",ten_biggest)]
monthly_close <- monthly_close[,c("date",ten_biggest)]
#Making daily, weekly and monthly volume for ten biggest coins
daily_volume <- spread(df[,c(1,4,10)],slug,volume)
weekly_volume <- spread(week_xts[,c(1,2,10)],slug,volume)
monthly_volume <- spread(month_xts[,c(1,2,10)],slug,volume)
daily_volume <- daily_volume[,c("date",ten_biggest)]
weekly_volume <- weekly_volume[,c("date",ten_biggest)]
monthly_volume <- monthly_volume[,c("date",ten_biggest)]
#Making daily,weekly and monthly market for ten biggest coins
daily_market <- spread(df[,c(1,4,11)],slug,market)
weekly_market <- spread(week_xts[,c(1,2,11)],slug,market)
monthly_market <- spread(month_xts[,c(1,2,11)],slug,market)
daily_market <- daily_market[,c("date",ten_biggest)]
weekly_market <- weekly_market[,c("date",ten_biggest)]
monthly_market <- monthly_market[,c("date",ten_biggest)]
#Doesn't work
weekly_open_returns <- periodReturn(weekly_open, period="daily", subset=NULL, type="arithmetic",leading="TRUE")
weekly_open_returns <- Return.calculate(weekly_open)
我想为每日，每月和每周频率的每种数据类型（开放，高，收盘，交易量，市场，close_ratio和价差）制作数据框。这些数据框的构造方式应使每列代表一种货币，行代表观察。但是，当我尝试使用每周和每月频率的扩展函数来执行此操作时，会出现错误。我该怎么办？
提前感谢您的帮助。祝你度过愉快的一天。
错误是：
> weekly_open_returns <- periodReturn(weekly_open, period="daily", subset=NULL, type="arithmetic",leading="TRUE")
Error in try.xts(x) :
Error in as.POSIXlt.character(x, tz, ...) : character string is not in a standard unambiguous format
> weekly_open_returns <- Return.calculate(weekly_open)
Error in checkData(prices, method = "xts") :
The data cannot be converted into a time series. If you are trying to pass in names from a data object with one column, you should use the form 'data[rows, columns, drop = FALSE]'. Rownames should have standard date formats, such as '1985-03-15'.
（来自评论:)
我想为每日，每月和每周频率的每种数据类型（开放，高，收盘，交易量，市场，close_ratio和价差）制作数据框。这些数据框的构造方式应使每列代表一种货币，行代表观察。但是，当我尝试使用每周和每月频率的扩展函数来执行此操作时，会出现错误。我该怎么办？提前感谢您的帮助。祝你愉快。 - MP PM 28分钟前
答案 0 :(得分：0)
在测试代码之后，错误非常明确：tidyr::spread适用于tibbles，但你拥有的是类[1] "xts" "zoo"。此外，可能由于xts步骤，它全部为character，这意味着week_xts中的所有数字都是字符串。我看到您认为自己在使用daily_xts（后spread）时所做的工作，但大部分数据都是NA，因为大多数货币直到数据末期才会出现。这是故意的吗？
部分问题在于您使用xts：它不会返回data.frame，而是返回一个向量或matrix，如下所示：
str(week_xts)
# An 'xts' object on 2013-04-27 17:00:00/2018-04-28 17:00:00 containing:
# Data: chr [1:4593, 1:10] "bitcoin" "litecoin" "bitcoin" "litecoin" "bitcoin" ...
# - attr(*, "dimnames")=List of 2
# ..$ : NULL
# ..$ : chr [1:10] "slug" "symbol" "name" "ranknow" ...
# Indexed by objects of class: [POSIXct,POSIXt] TZ:
# xts Attributes:
# NULL
这表示一切都是character。所以，如果你这样做
head(week_xts[,1:5])
# slug symbol name ranknow open
# 2013-04-27 17:00:00 "bitcoin" "BTC" "Bitcoin" " 1" " 135.300000"
# 2013-04-27 17:00:00 "litecoin" "LTC" "Litecoin" " 7" " 4.300000"
# 2013-05-04 17:00:00 "bitcoin" "BTC" "Bitcoin" " 1" " 116.380000"
# 2013-05-04 17:00:00 "litecoin" "LTC" "Litecoin" " 7" " 3.780000"
# 2013-05-11 17:00:00 "bitcoin" "BTC" "Bitcoin" " 1" " 113.200000"
# 2013-05-11 17:00:00 "litecoin" "LTC" "Litecoin" " 7" " 3.400000"
你会看到你想要的数字实际上是字符串。
我建议yourfunction或许xts::xts不应该在你spread事之前做yourfunction <- function(df, frequency = NULL) {
# ...
return(data)
}
week_xts <- yourfunction(df, frequency = "week")
str(week_xts)
# 'data.frame': 4593 obs. of 11 variables:
# $ date : Date, format: "2013-04-28" "2013-04-28" ...
# $ slug : chr "bitcoin" "litecoin" "bitcoin" "litecoin" ...
# $ symbol : chr "BTC" "LTC" "BTC" "LTC" ...
# $ name : chr "Bitcoin" "Litecoin" "Bitcoin" "Litecoin" ...
# $ ranknow: num 1 7 1 7 1 7 1 7 1 7 ...
# $ open : num 135.3 4.3 116.38 3.78 113.2 ...
# $ high : num 147.49 4.57 125.6 4.04 122 ...
# $ low : num 107.72 3.52 79.1 2.4 103.5 ...
# $ close : num 116.99 3.8 113.57 3.41 114.22 ...
# $ volume : num 0 0 0 0 0 0 0 0 0 0 ...
# $ market : num 1542820000 73901200 1219450000 57196300 1242760000 ...
。代替：
as.tbl(week_xts) %>%
select(date, slug, open) %>%
spread(slug, open) %>%
tail()
# # A tibble: 6 x 51
# date `0x` aelf aeternity `binance-coin` bitcoin `bitcoin-cash`
# <date> <dbl> <dbl> <dbl> <dbl> <dbl> <dbl>
# 1 2018-03-25 0.583 0.633 1.75 10.0 8939. 1033.
# 2 2018-04-01 0.635 0.625 1.64 11.7 7979. 862.
# 3 2018-04-08 0.521 0.519 1.43 12.2 6849. 649.
# 4 2018-04-15 0.597 0.895 1.47 12.4 6955. 666.
# 5 2018-04-22 0.923 1.11 1.86 12.3 8159. 891.
# 6 2018-04-29 1.01 1.02 2.3 13.4 8867. 1290.
# # ... with 44 more variables: `bitcoin-diamond` <dbl>, `bitcoin-gold` <dbl>,
# # `bitcoin-private` <dbl>, bitshares <dbl>, `bytecoin-bcn` <dbl>,
# # bytom <dbl>, cardano <dbl>, dash <dbl>, decred <dbl>, digixdao <dbl>,
# # dogecoin <dbl>, eos <dbl>, ethereum <dbl>, `ethereum-classic` <dbl>,
# # icon <dbl>, iota <dbl>, lisk <dbl>, litecoin <dbl>, loopring <dbl>,
# # maker <dbl>, mixin <dbl>, monero <dbl>, nano <dbl>, nem <dbl>, neo <dbl>,
# # omisego <dbl>, ontology <dbl>, populous <dbl>, qtum <dbl>, rchain <dbl>,
# # ripple <dbl>, siacoin <dbl>, status <dbl>, steem <dbl>, stellar <dbl>,
# # stratis <dbl>, tether <dbl>, tron <dbl>, vechain <dbl>, verge <dbl>,
# # wanchain <dbl>, waves <dbl>, zcash <dbl>, zilliqa <dbl>
从那里：
tail（我显示select(...)因为大多数货币的大部分早期日期都是空的。）
附注：我建议您使用week_xts[,c(1,5)]和列名而不是索引;您使用了匹配date和open的{{1}}，是的，但是如果不查看数据则不是很清楚。此外，通过跳过xts转换，现在c(1,2,6)可以捕获日期，slug和open。
我想知道这是否应该考虑使用xts::xts，尽管数据中包含slug，您仍可能将所有数字转换为character。
最后一个日期的20个条目，如果其他人想要快速了解这个：
> dput(head(filter(df, date==tail(date,1)),n=20))
structure(list(slug = c("bitcoin", "ethereum", "ripple", "bitcoin-cash",
"eos", "cardano", "litecoin", "stellar", "tron", "neo", "iota",
"dash", "monero", "nem", "tether", "vechain", "ethereum-classic",
"qtum", "omisego", "icon"), symbol = c("BTC", "ETH", "XRP", "BCH",
"EOS", "ADA", "LTC", "XLM", "TRX", "NEO", "MIOTA", "DASH", "XMR",
"XEM", "USDT", "VEN", "ETC", "QTUM", "OMG", "ICX"), name = c("Bitcoin",
"Ethereum", "Ripple", "Bitcoin Cash", "EOS", "Cardano", "Litecoin",
"Stellar", "TRON", "NEO", "IOTA", "Dash", "Monero", "NEM", "Tether",
"VeChain", "Ethereum Classic", "Qtum", "OmiseGO", "ICON"), date = structure(c(17651,
17651, 17651, 17651, 17651, 17651, 17651, 17651, 17651, 17651,
17651, 17651, 17651, 17651, 17651, 17651, 17651, 17651, 17651,
17651), class = "Date"), ranknow = c(1, 2, 3, 4, 5, 6, 7, 8,
9, 10, 11, 12, 13, 14, 15, 16, 17, 18, 19, 20), open = c(9426.11,
689.76, 0.871404, 1440.96, 21.64, 0.364297, 153.65, 0.458688,
0.089656, 92.15, 2.04, 498.44, 256.35, 0.42658, 0.997553, 4.83,
21.84, 23.15, 18.05, 4.69), high = c(9477.14, 694.44, 0.876788,
1440.96, 21.64, 0.364297, 154.08, 0.460128, 0.101197, 94.76,
2.05, 499.43, 259.07, 0.42658, 1.01, 4.83, 22.95, 25.6, 18.06,
4.69), low = c(9166.81, 666.12, 0.831208, 1339.36, 16.86, 0.336625,
147.87, 0.417446, 0.086102, 82.86, 1.94, 472.02, 239.08, 0.403889,
0.992921, 4.37, 21.42, 22.12, 16.81, 4.26), close = c(9240.55,
669.92, 0.837938, 1350.05, 17.58, 0.343318, 148.48, 0.424659,
0.093777, 84.27, 1.96, 472.77, 242.46, 0.40888, 0.998919, 4.49,
21.68, 22.64, 16.95, 4.34), volume = c(8673920000, 2853100000,
575364000, 753114000, 4073370000, 298712000, 341397000, 81453300,
1749640000, 377385000, 61762500, 118497000, 103574000, 25523800,
4498440000, 109782000, 351049000, 414455000, 68967800, 94284100
), market = c(160302000000, 68376400000, 34112200000, 24642000000,
17849100000, 9445160000, 8651810000, 8518430000, 5894710000,
5990010000, 5676530000, 4006280000, 4096790000, 3839220000, 2411230000,
2539730000, 2215850000, 2050300000, 1841910000, 1815430000),
close_ratio = c(0.2376, 0.1342, 0.1477, 0.1052, 0.1506, 0.2419,
0.0982, 0.169, 0.5084, 0.1185, 0.1818, 0.0274, 0.1691, 0.22,
0.3512, 0.2609, 0.1699, 0.1494, 0.112, 0.186), spread = c(310.33,
28.32, 0.05, 101.6, 4.78, 0.03, 6.21, 0.04, 0.02, 11.9, 0.11,
27.41, 19.99, 0.02, 0.02, 0.46, 1.53, 3.48, 1.25, 0.43)), .Names = c("slug",
"symbol", "name", "date", "ranknow", "open", "high", "low", "close",
"volume", "market", "close_ratio", "spread"), row.names = c(NA,
20L), class = "data.frame")
答案 1 :(得分：0)
感谢您的宝贵意见。我编辑了代码（你可以在第一篇文章中找到新的代码），因为当我想要计算回报时，我还有另一个挑战。而且，如果有人会有一些提示，我会非常高兴。也许直到星期二，我将用退货计算解决我的问题。